{
  "best_metric": 0.10501641035079956,
  "best_model_checkpoint": "./vit-base-lcdoctypev1_session3/checkpoint-335",
  "epoch": 10.0,
  "eval_steps": 5,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.3158922493457794,
      "eval_runtime": 9.7781,
      "eval_samples_per_second": 12.375,
      "eval_steps_per_second": 1.636,
      "step": 5
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.890113353729248,
      "learning_rate": 0.00019666666666666666,
      "loss": 0.1798,
      "step": 10
    },
    {
      "epoch": 0.17,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.22616656124591827,
      "eval_runtime": 8.5759,
      "eval_samples_per_second": 14.109,
      "eval_steps_per_second": 1.866,
      "step": 10
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.7768595041322314,
      "eval_loss": 0.9910252690315247,
      "eval_runtime": 9.2477,
      "eval_samples_per_second": 13.084,
      "eval_steps_per_second": 1.73,
      "step": 15
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.934413909912109,
      "learning_rate": 0.0001936666666666667,
      "loss": 0.3815,
      "step": 20
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.9008264462809917,
      "eval_loss": 0.3035498857498169,
      "eval_runtime": 8.7721,
      "eval_samples_per_second": 13.794,
      "eval_steps_per_second": 1.824,
      "step": 20
    },
    {
      "epoch": 0.42,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.21773552894592285,
      "eval_runtime": 9.0833,
      "eval_samples_per_second": 13.321,
      "eval_steps_per_second": 1.761,
      "step": 25
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3238607347011566,
      "learning_rate": 0.00019033333333333334,
      "loss": 0.1429,
      "step": 30
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.8842975206611571,
      "eval_loss": 0.4908874034881592,
      "eval_runtime": 8.456,
      "eval_samples_per_second": 14.309,
      "eval_steps_per_second": 1.892,
      "step": 30
    },
    {
      "epoch": 0.58,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.3096350133419037,
      "eval_runtime": 10.0423,
      "eval_samples_per_second": 12.049,
      "eval_steps_per_second": 1.593,
      "step": 35
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.55199670791626,
      "learning_rate": 0.00018700000000000002,
      "loss": 0.2424,
      "step": 40
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.32702526450157166,
      "eval_runtime": 8.357,
      "eval_samples_per_second": 14.479,
      "eval_steps_per_second": 1.915,
      "step": 40
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.2554916441440582,
      "eval_runtime": 9.2174,
      "eval_samples_per_second": 13.127,
      "eval_steps_per_second": 1.736,
      "step": 45
    },
    {
      "epoch": 0.83,
      "grad_norm": 14.463412284851074,
      "learning_rate": 0.00018366666666666667,
      "loss": 0.1172,
      "step": 50
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.2309066504240036,
      "eval_runtime": 9.1271,
      "eval_samples_per_second": 13.257,
      "eval_steps_per_second": 1.753,
      "step": 50
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.2952496409416199,
      "eval_runtime": 9.6618,
      "eval_samples_per_second": 12.523,
      "eval_steps_per_second": 1.656,
      "step": 55
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9041020274162292,
      "learning_rate": 0.00018033333333333334,
      "loss": 0.1185,
      "step": 60
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.295705646276474,
      "eval_runtime": 8.6323,
      "eval_samples_per_second": 14.017,
      "eval_steps_per_second": 1.854,
      "step": 60
    },
    {
      "epoch": 1.08,
      "eval_accuracy": 0.8925619834710744,
      "eval_loss": 0.3724129796028137,
      "eval_runtime": 8.5924,
      "eval_samples_per_second": 14.082,
      "eval_steps_per_second": 1.862,
      "step": 65
    },
    {
      "epoch": 1.17,
      "grad_norm": 24.987598419189453,
      "learning_rate": 0.00017700000000000002,
      "loss": 0.1594,
      "step": 70
    },
    {
      "epoch": 1.17,
      "eval_accuracy": 0.8842975206611571,
      "eval_loss": 0.4216250777244568,
      "eval_runtime": 9.0834,
      "eval_samples_per_second": 13.321,
      "eval_steps_per_second": 1.761,
      "step": 70
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.347516268491745,
      "eval_runtime": 8.9447,
      "eval_samples_per_second": 13.528,
      "eval_steps_per_second": 1.789,
      "step": 75
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3783435523509979,
      "learning_rate": 0.00017366666666666667,
      "loss": 0.1231,
      "step": 80
    },
    {
      "epoch": 1.33,
      "eval_accuracy": 0.8925619834710744,
      "eval_loss": 0.323406845331192,
      "eval_runtime": 9.122,
      "eval_samples_per_second": 13.265,
      "eval_steps_per_second": 1.754,
      "step": 80
    },
    {
      "epoch": 1.42,
      "eval_accuracy": 0.8842975206611571,
      "eval_loss": 0.4309641718864441,
      "eval_runtime": 8.8847,
      "eval_samples_per_second": 13.619,
      "eval_steps_per_second": 1.801,
      "step": 85
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.4656147956848145,
      "learning_rate": 0.00017033333333333334,
      "loss": 0.0875,
      "step": 90
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.3598105013370514,
      "eval_runtime": 9.086,
      "eval_samples_per_second": 13.317,
      "eval_steps_per_second": 1.761,
      "step": 90
    },
    {
      "epoch": 1.58,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.3038037419319153,
      "eval_runtime": 8.8538,
      "eval_samples_per_second": 13.666,
      "eval_steps_per_second": 1.807,
      "step": 95
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.0762423649430275,
      "learning_rate": 0.000167,
      "loss": 0.0897,
      "step": 100
    },
    {
      "epoch": 1.67,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.25987809896469116,
      "eval_runtime": 8.7997,
      "eval_samples_per_second": 13.751,
      "eval_steps_per_second": 1.818,
      "step": 100
    },
    {
      "epoch": 1.75,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.1683536171913147,
      "eval_runtime": 8.0037,
      "eval_samples_per_second": 15.118,
      "eval_steps_per_second": 1.999,
      "step": 105
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.4948017597198486,
      "learning_rate": 0.00016366666666666667,
      "loss": 0.1797,
      "step": 110
    },
    {
      "epoch": 1.83,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.1412244588136673,
      "eval_runtime": 8.3997,
      "eval_samples_per_second": 14.405,
      "eval_steps_per_second": 1.905,
      "step": 110
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.14531350135803223,
      "eval_runtime": 8.6407,
      "eval_samples_per_second": 14.003,
      "eval_steps_per_second": 1.852,
      "step": 115
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.408501148223877,
      "learning_rate": 0.00016033333333333335,
      "loss": 0.1178,
      "step": 120
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8925619834710744,
      "eval_loss": 0.3830728530883789,
      "eval_runtime": 9.0064,
      "eval_samples_per_second": 13.435,
      "eval_steps_per_second": 1.777,
      "step": 120
    },
    {
      "epoch": 2.08,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.3321413993835449,
      "eval_runtime": 8.8462,
      "eval_samples_per_second": 13.678,
      "eval_steps_per_second": 1.809,
      "step": 125
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.7101590633392334,
      "learning_rate": 0.00015700000000000002,
      "loss": 0.1969,
      "step": 130
    },
    {
      "epoch": 2.17,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.25461918115615845,
      "eval_runtime": 8.969,
      "eval_samples_per_second": 13.491,
      "eval_steps_per_second": 1.784,
      "step": 130
    },
    {
      "epoch": 2.25,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.18391890823841095,
      "eval_runtime": 8.644,
      "eval_samples_per_second": 13.998,
      "eval_steps_per_second": 1.851,
      "step": 135
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.9350224733352661,
      "learning_rate": 0.00015366666666666667,
      "loss": 0.0362,
      "step": 140
    },
    {
      "epoch": 2.33,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.20266053080558777,
      "eval_runtime": 8.5546,
      "eval_samples_per_second": 14.144,
      "eval_steps_per_second": 1.87,
      "step": 140
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.9090909090909091,
      "eval_loss": 0.28766530752182007,
      "eval_runtime": 8.0124,
      "eval_samples_per_second": 15.102,
      "eval_steps_per_second": 1.997,
      "step": 145
    },
    {
      "epoch": 2.5,
      "grad_norm": 8.35146427154541,
      "learning_rate": 0.00015033333333333335,
      "loss": 0.1047,
      "step": 150
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.8925619834710744,
      "eval_loss": 0.4503512978553772,
      "eval_runtime": 8.454,
      "eval_samples_per_second": 14.313,
      "eval_steps_per_second": 1.893,
      "step": 150
    },
    {
      "epoch": 2.58,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.1810603141784668,
      "eval_runtime": 9.2694,
      "eval_samples_per_second": 13.054,
      "eval_steps_per_second": 1.726,
      "step": 155
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.6059785485267639,
      "learning_rate": 0.000147,
      "loss": 0.1232,
      "step": 160
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.21074515581130981,
      "eval_runtime": 8.8489,
      "eval_samples_per_second": 13.674,
      "eval_steps_per_second": 1.808,
      "step": 160
    },
    {
      "epoch": 2.75,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.20863419771194458,
      "eval_runtime": 8.9684,
      "eval_samples_per_second": 13.492,
      "eval_steps_per_second": 1.784,
      "step": 165
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.04269757494330406,
      "learning_rate": 0.00014366666666666667,
      "loss": 0.0611,
      "step": 170
    },
    {
      "epoch": 2.83,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.2971450686454773,
      "eval_runtime": 9.1231,
      "eval_samples_per_second": 13.263,
      "eval_steps_per_second": 1.754,
      "step": 170
    },
    {
      "epoch": 2.92,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.2731765806674957,
      "eval_runtime": 8.9974,
      "eval_samples_per_second": 13.448,
      "eval_steps_per_second": 1.778,
      "step": 175
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9442762732505798,
      "learning_rate": 0.00014033333333333335,
      "loss": 0.0815,
      "step": 180
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.16794723272323608,
      "eval_runtime": 8.9455,
      "eval_samples_per_second": 13.526,
      "eval_steps_per_second": 1.789,
      "step": 180
    },
    {
      "epoch": 3.08,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.24155646562576294,
      "eval_runtime": 8.3613,
      "eval_samples_per_second": 14.471,
      "eval_steps_per_second": 1.914,
      "step": 185
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.08608078956604,
      "learning_rate": 0.00013700000000000002,
      "loss": 0.0469,
      "step": 190
    },
    {
      "epoch": 3.17,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.29269692301750183,
      "eval_runtime": 8.9407,
      "eval_samples_per_second": 13.534,
      "eval_steps_per_second": 1.79,
      "step": 190
    },
    {
      "epoch": 3.25,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.28314581513404846,
      "eval_runtime": 8.7918,
      "eval_samples_per_second": 13.763,
      "eval_steps_per_second": 1.82,
      "step": 195
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.8798501491546631,
      "learning_rate": 0.00013366666666666667,
      "loss": 0.0443,
      "step": 200
    },
    {
      "epoch": 3.33,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.2744951546192169,
      "eval_runtime": 8.612,
      "eval_samples_per_second": 14.05,
      "eval_steps_per_second": 1.858,
      "step": 200
    },
    {
      "epoch": 3.42,
      "eval_accuracy": 0.8925619834710744,
      "eval_loss": 0.4193201959133148,
      "eval_runtime": 8.9147,
      "eval_samples_per_second": 13.573,
      "eval_steps_per_second": 1.795,
      "step": 205
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.03738969564437866,
      "learning_rate": 0.00013033333333333332,
      "loss": 0.0823,
      "step": 210
    },
    {
      "epoch": 3.5,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.3746081292629242,
      "eval_runtime": 8.5854,
      "eval_samples_per_second": 14.094,
      "eval_steps_per_second": 1.864,
      "step": 210
    },
    {
      "epoch": 3.58,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.30296453833580017,
      "eval_runtime": 8.8651,
      "eval_samples_per_second": 13.649,
      "eval_steps_per_second": 1.805,
      "step": 215
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.03203318268060684,
      "learning_rate": 0.000127,
      "loss": 0.0101,
      "step": 220
    },
    {
      "epoch": 3.67,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.21464580297470093,
      "eval_runtime": 8.8029,
      "eval_samples_per_second": 13.745,
      "eval_steps_per_second": 1.818,
      "step": 220
    },
    {
      "epoch": 3.75,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.2514008581638336,
      "eval_runtime": 9.2073,
      "eval_samples_per_second": 13.142,
      "eval_steps_per_second": 1.738,
      "step": 225
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.04610053077340126,
      "learning_rate": 0.00012366666666666667,
      "loss": 0.16,
      "step": 230
    },
    {
      "epoch": 3.83,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.25517505407333374,
      "eval_runtime": 8.8885,
      "eval_samples_per_second": 13.613,
      "eval_steps_per_second": 1.8,
      "step": 230
    },
    {
      "epoch": 3.92,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.22389596700668335,
      "eval_runtime": 8.5203,
      "eval_samples_per_second": 14.201,
      "eval_steps_per_second": 1.878,
      "step": 235
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.5699214935302734,
      "learning_rate": 0.00012033333333333335,
      "loss": 0.1687,
      "step": 240
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.25712552666664124,
      "eval_runtime": 8.7329,
      "eval_samples_per_second": 13.856,
      "eval_steps_per_second": 1.832,
      "step": 240
    },
    {
      "epoch": 4.08,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13568438589572906,
      "eval_runtime": 8.642,
      "eval_samples_per_second": 14.001,
      "eval_steps_per_second": 1.851,
      "step": 245
    },
    {
      "epoch": 4.17,
      "grad_norm": 2.446256637573242,
      "learning_rate": 0.000117,
      "loss": 0.0758,
      "step": 250
    },
    {
      "epoch": 4.17,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.17341962456703186,
      "eval_runtime": 8.0872,
      "eval_samples_per_second": 14.962,
      "eval_steps_per_second": 1.978,
      "step": 250
    },
    {
      "epoch": 4.25,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.11965644359588623,
      "eval_runtime": 9.0168,
      "eval_samples_per_second": 13.419,
      "eval_steps_per_second": 1.774,
      "step": 255
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.048665851354599,
      "learning_rate": 0.00011366666666666667,
      "loss": 0.042,
      "step": 260
    },
    {
      "epoch": 4.33,
      "eval_accuracy": 0.9421487603305785,
      "eval_loss": 0.23387375473976135,
      "eval_runtime": 8.7974,
      "eval_samples_per_second": 13.754,
      "eval_steps_per_second": 1.819,
      "step": 260
    },
    {
      "epoch": 4.42,
      "eval_accuracy": 0.9173553719008265,
      "eval_loss": 0.2923980951309204,
      "eval_runtime": 8.7236,
      "eval_samples_per_second": 13.87,
      "eval_steps_per_second": 1.834,
      "step": 265
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.07859846204519272,
      "learning_rate": 0.00011033333333333334,
      "loss": 0.0114,
      "step": 270
    },
    {
      "epoch": 4.5,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.23183898627758026,
      "eval_runtime": 8.7284,
      "eval_samples_per_second": 13.863,
      "eval_steps_per_second": 1.833,
      "step": 270
    },
    {
      "epoch": 4.58,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.17654092609882355,
      "eval_runtime": 8.8357,
      "eval_samples_per_second": 13.694,
      "eval_steps_per_second": 1.811,
      "step": 275
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.266812801361084,
      "learning_rate": 0.00010700000000000001,
      "loss": 0.0197,
      "step": 280
    },
    {
      "epoch": 4.67,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12631382048130035,
      "eval_runtime": 7.9448,
      "eval_samples_per_second": 15.23,
      "eval_steps_per_second": 2.014,
      "step": 280
    },
    {
      "epoch": 4.75,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12528358399868011,
      "eval_runtime": 8.1841,
      "eval_samples_per_second": 14.785,
      "eval_steps_per_second": 1.955,
      "step": 285
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.8625770807266235,
      "learning_rate": 0.00010366666666666666,
      "loss": 0.0283,
      "step": 290
    },
    {
      "epoch": 4.83,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1239379420876503,
      "eval_runtime": 9.1766,
      "eval_samples_per_second": 13.186,
      "eval_steps_per_second": 1.744,
      "step": 290
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12782499194145203,
      "eval_runtime": 8.9807,
      "eval_samples_per_second": 13.473,
      "eval_steps_per_second": 1.782,
      "step": 295
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.08663175255060196,
      "learning_rate": 0.00010033333333333335,
      "loss": 0.1115,
      "step": 300
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.2527827322483063,
      "eval_runtime": 8.5733,
      "eval_samples_per_second": 14.114,
      "eval_steps_per_second": 1.866,
      "step": 300
    },
    {
      "epoch": 5.08,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.3164093792438507,
      "eval_runtime": 8.5647,
      "eval_samples_per_second": 14.128,
      "eval_steps_per_second": 1.868,
      "step": 305
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.11408742517232895,
      "learning_rate": 9.7e-05,
      "loss": 0.0404,
      "step": 310
    },
    {
      "epoch": 5.17,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.2841833829879761,
      "eval_runtime": 8.7395,
      "eval_samples_per_second": 13.845,
      "eval_steps_per_second": 1.831,
      "step": 310
    },
    {
      "epoch": 5.25,
      "eval_accuracy": 0.9504132231404959,
      "eval_loss": 0.17133790254592896,
      "eval_runtime": 8.8895,
      "eval_samples_per_second": 13.612,
      "eval_steps_per_second": 1.8,
      "step": 315
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.08482904732227325,
      "learning_rate": 9.366666666666668e-05,
      "loss": 0.0719,
      "step": 320
    },
    {
      "epoch": 5.33,
      "eval_accuracy": 0.9338842975206612,
      "eval_loss": 0.18959270417690277,
      "eval_runtime": 8.641,
      "eval_samples_per_second": 14.003,
      "eval_steps_per_second": 1.852,
      "step": 320
    },
    {
      "epoch": 5.42,
      "eval_accuracy": 0.9256198347107438,
      "eval_loss": 0.18550463020801544,
      "eval_runtime": 8.219,
      "eval_samples_per_second": 14.722,
      "eval_steps_per_second": 1.947,
      "step": 325
    },
    {
      "epoch": 5.5,
      "grad_norm": 4.354619026184082,
      "learning_rate": 9.033333333333334e-05,
      "loss": 0.0435,
      "step": 330
    },
    {
      "epoch": 5.5,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.15409986674785614,
      "eval_runtime": 8.5474,
      "eval_samples_per_second": 14.156,
      "eval_steps_per_second": 1.872,
      "step": 330
    },
    {
      "epoch": 5.58,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.10501641035079956,
      "eval_runtime": 8.6962,
      "eval_samples_per_second": 13.914,
      "eval_steps_per_second": 1.84,
      "step": 335
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.05306649208068848,
      "learning_rate": 8.7e-05,
      "loss": 0.0129,
      "step": 340
    },
    {
      "epoch": 5.67,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.10632016509771347,
      "eval_runtime": 8.7849,
      "eval_samples_per_second": 13.774,
      "eval_steps_per_second": 1.821,
      "step": 340
    },
    {
      "epoch": 5.75,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.11378511786460876,
      "eval_runtime": 8.0973,
      "eval_samples_per_second": 14.943,
      "eval_steps_per_second": 1.976,
      "step": 345
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.02987060882151127,
      "learning_rate": 8.366666666666668e-05,
      "loss": 0.0222,
      "step": 350
    },
    {
      "epoch": 5.83,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.11444854736328125,
      "eval_runtime": 8.7513,
      "eval_samples_per_second": 13.827,
      "eval_steps_per_second": 1.828,
      "step": 350
    },
    {
      "epoch": 5.92,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12378235161304474,
      "eval_runtime": 9.2818,
      "eval_samples_per_second": 13.036,
      "eval_steps_per_second": 1.724,
      "step": 355
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02307078428566456,
      "learning_rate": 8.033333333333334e-05,
      "loss": 0.0431,
      "step": 360
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.1342514306306839,
      "eval_runtime": 7.9783,
      "eval_samples_per_second": 15.166,
      "eval_steps_per_second": 2.005,
      "step": 360
    },
    {
      "epoch": 6.08,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.144140362739563,
      "eval_runtime": 8.763,
      "eval_samples_per_second": 13.808,
      "eval_steps_per_second": 1.826,
      "step": 365
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.016979066655039787,
      "learning_rate": 7.7e-05,
      "loss": 0.0064,
      "step": 370
    },
    {
      "epoch": 6.17,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1470753401517868,
      "eval_runtime": 8.9895,
      "eval_samples_per_second": 13.46,
      "eval_steps_per_second": 1.78,
      "step": 370
    },
    {
      "epoch": 6.25,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.1360587477684021,
      "eval_runtime": 8.9513,
      "eval_samples_per_second": 13.518,
      "eval_steps_per_second": 1.787,
      "step": 375
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.022434255108237267,
      "learning_rate": 7.366666666666668e-05,
      "loss": 0.0576,
      "step": 380
    },
    {
      "epoch": 6.33,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13161548972129822,
      "eval_runtime": 8.7711,
      "eval_samples_per_second": 13.795,
      "eval_steps_per_second": 1.824,
      "step": 380
    },
    {
      "epoch": 6.42,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12319940328598022,
      "eval_runtime": 8.3275,
      "eval_samples_per_second": 14.53,
      "eval_steps_per_second": 1.921,
      "step": 385
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.2556796371936798,
      "learning_rate": 7.033333333333334e-05,
      "loss": 0.0298,
      "step": 390
    },
    {
      "epoch": 6.5,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1254538893699646,
      "eval_runtime": 8.6034,
      "eval_samples_per_second": 14.064,
      "eval_steps_per_second": 1.86,
      "step": 390
    },
    {
      "epoch": 6.58,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13591305911540985,
      "eval_runtime": 8.8244,
      "eval_samples_per_second": 13.712,
      "eval_steps_per_second": 1.813,
      "step": 395
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.03436155617237091,
      "learning_rate": 6.7e-05,
      "loss": 0.0097,
      "step": 400
    },
    {
      "epoch": 6.67,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1434980034828186,
      "eval_runtime": 9.1677,
      "eval_samples_per_second": 13.199,
      "eval_steps_per_second": 1.745,
      "step": 400
    },
    {
      "epoch": 6.75,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.14506025612354279,
      "eval_runtime": 8.7551,
      "eval_samples_per_second": 13.82,
      "eval_steps_per_second": 1.827,
      "step": 405
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.019292179495096207,
      "learning_rate": 6.366666666666668e-05,
      "loss": 0.0153,
      "step": 410
    },
    {
      "epoch": 6.83,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.14391547441482544,
      "eval_runtime": 8.6401,
      "eval_samples_per_second": 14.004,
      "eval_steps_per_second": 1.852,
      "step": 410
    },
    {
      "epoch": 6.92,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.1352916657924652,
      "eval_runtime": 8.9781,
      "eval_samples_per_second": 13.477,
      "eval_steps_per_second": 1.782,
      "step": 415
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1764516532421112,
      "learning_rate": 6.033333333333334e-05,
      "loss": 0.0406,
      "step": 420
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13157208263874054,
      "eval_runtime": 8.4013,
      "eval_samples_per_second": 14.402,
      "eval_steps_per_second": 1.904,
      "step": 420
    },
    {
      "epoch": 7.08,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13093091547489166,
      "eval_runtime": 9.2645,
      "eval_samples_per_second": 13.061,
      "eval_steps_per_second": 1.727,
      "step": 425
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.025451194494962692,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 0.0154,
      "step": 430
    },
    {
      "epoch": 7.17,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13050581514835358,
      "eval_runtime": 8.9669,
      "eval_samples_per_second": 13.494,
      "eval_steps_per_second": 1.784,
      "step": 430
    },
    {
      "epoch": 7.25,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13096679747104645,
      "eval_runtime": 8.8412,
      "eval_samples_per_second": 13.686,
      "eval_steps_per_second": 1.81,
      "step": 435
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.01770775578916073,
      "learning_rate": 5.3666666666666666e-05,
      "loss": 0.0209,
      "step": 440
    },
    {
      "epoch": 7.33,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.13012637197971344,
      "eval_runtime": 8.8578,
      "eval_samples_per_second": 13.66,
      "eval_steps_per_second": 1.806,
      "step": 440
    },
    {
      "epoch": 7.42,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.145931214094162,
      "eval_runtime": 8.6371,
      "eval_samples_per_second": 14.009,
      "eval_steps_per_second": 1.852,
      "step": 445
    },
    {
      "epoch": 7.5,
      "grad_norm": 17.443572998046875,
      "learning_rate": 5.0333333333333335e-05,
      "loss": 0.0298,
      "step": 450
    },
    {
      "epoch": 7.5,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.16629938781261444,
      "eval_runtime": 8.2752,
      "eval_samples_per_second": 14.622,
      "eval_steps_per_second": 1.933,
      "step": 450
    },
    {
      "epoch": 7.58,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.15594066679477692,
      "eval_runtime": 8.806,
      "eval_samples_per_second": 13.741,
      "eval_steps_per_second": 1.817,
      "step": 455
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.030463455244898796,
      "learning_rate": 4.7e-05,
      "loss": 0.0052,
      "step": 460
    },
    {
      "epoch": 7.67,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.15159618854522705,
      "eval_runtime": 8.9863,
      "eval_samples_per_second": 13.465,
      "eval_steps_per_second": 1.78,
      "step": 460
    },
    {
      "epoch": 7.75,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.13964863121509552,
      "eval_runtime": 8.4469,
      "eval_samples_per_second": 14.325,
      "eval_steps_per_second": 1.894,
      "step": 465
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.11391649395227432,
      "learning_rate": 4.3666666666666666e-05,
      "loss": 0.0172,
      "step": 470
    },
    {
      "epoch": 7.83,
      "eval_accuracy": 0.9586776859504132,
      "eval_loss": 0.13303633034229279,
      "eval_runtime": 8.9949,
      "eval_samples_per_second": 13.452,
      "eval_steps_per_second": 1.779,
      "step": 470
    },
    {
      "epoch": 7.92,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12355250120162964,
      "eval_runtime": 8.1109,
      "eval_samples_per_second": 14.918,
      "eval_steps_per_second": 1.973,
      "step": 475
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.059335533529520035,
      "learning_rate": 4.0333333333333336e-05,
      "loss": 0.0348,
      "step": 480
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12098132073879242,
      "eval_runtime": 8.8837,
      "eval_samples_per_second": 13.621,
      "eval_steps_per_second": 1.801,
      "step": 480
    },
    {
      "epoch": 8.08,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.11751802265644073,
      "eval_runtime": 9.8045,
      "eval_samples_per_second": 12.341,
      "eval_steps_per_second": 1.632,
      "step": 485
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.019469719380140305,
      "learning_rate": 3.7e-05,
      "loss": 0.0068,
      "step": 490
    },
    {
      "epoch": 8.17,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.1185019463300705,
      "eval_runtime": 8.8757,
      "eval_samples_per_second": 13.633,
      "eval_steps_per_second": 1.803,
      "step": 490
    },
    {
      "epoch": 8.25,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12291049212217331,
      "eval_runtime": 8.9999,
      "eval_samples_per_second": 13.445,
      "eval_steps_per_second": 1.778,
      "step": 495
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.029128719121217728,
      "learning_rate": 3.366666666666667e-05,
      "loss": 0.0305,
      "step": 500
    },
    {
      "epoch": 8.33,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12297818809747696,
      "eval_runtime": 9.0782,
      "eval_samples_per_second": 13.329,
      "eval_steps_per_second": 1.762,
      "step": 500
    },
    {
      "epoch": 8.42,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12048203498125076,
      "eval_runtime": 8.2377,
      "eval_samples_per_second": 14.688,
      "eval_steps_per_second": 1.942,
      "step": 505
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.01591232791543007,
      "learning_rate": 3.0333333333333337e-05,
      "loss": 0.0154,
      "step": 510
    },
    {
      "epoch": 8.5,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.11965296417474747,
      "eval_runtime": 9.3646,
      "eval_samples_per_second": 12.921,
      "eval_steps_per_second": 1.709,
      "step": 510
    },
    {
      "epoch": 8.58,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12173377722501755,
      "eval_runtime": 9.3963,
      "eval_samples_per_second": 12.877,
      "eval_steps_per_second": 1.703,
      "step": 515
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.02258380502462387,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.0177,
      "step": 520
    },
    {
      "epoch": 8.67,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12387024611234665,
      "eval_runtime": 9.1961,
      "eval_samples_per_second": 13.158,
      "eval_steps_per_second": 1.74,
      "step": 520
    },
    {
      "epoch": 8.75,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12437601387500763,
      "eval_runtime": 9.196,
      "eval_samples_per_second": 13.158,
      "eval_steps_per_second": 1.74,
      "step": 525
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.020360412076115608,
      "learning_rate": 2.3666666666666668e-05,
      "loss": 0.0123,
      "step": 530
    },
    {
      "epoch": 8.83,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.12708893418312073,
      "eval_runtime": 8.8063,
      "eval_samples_per_second": 13.74,
      "eval_steps_per_second": 1.817,
      "step": 530
    },
    {
      "epoch": 8.92,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1300380975008011,
      "eval_runtime": 8.9605,
      "eval_samples_per_second": 13.504,
      "eval_steps_per_second": 1.786,
      "step": 535
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.6857224702835083,
      "learning_rate": 2.0333333333333334e-05,
      "loss": 0.0154,
      "step": 540
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13137011229991913,
      "eval_runtime": 8.8274,
      "eval_samples_per_second": 13.707,
      "eval_steps_per_second": 1.813,
      "step": 540
    },
    {
      "epoch": 9.08,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1295723170042038,
      "eval_runtime": 8.8666,
      "eval_samples_per_second": 13.647,
      "eval_steps_per_second": 1.805,
      "step": 545
    },
    {
      "epoch": 9.17,
      "grad_norm": 0.4675326943397522,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.0331,
      "step": 550
    },
    {
      "epoch": 9.17,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12511087954044342,
      "eval_runtime": 9.3211,
      "eval_samples_per_second": 12.981,
      "eval_steps_per_second": 1.717,
      "step": 550
    },
    {
      "epoch": 9.25,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12691423296928406,
      "eval_runtime": 9.1389,
      "eval_samples_per_second": 13.24,
      "eval_steps_per_second": 1.751,
      "step": 555
    },
    {
      "epoch": 9.33,
      "grad_norm": 0.02136796899139881,
      "learning_rate": 1.3666666666666666e-05,
      "loss": 0.0196,
      "step": 560
    },
    {
      "epoch": 9.33,
      "eval_accuracy": 0.9752066115702479,
      "eval_loss": 0.12836582958698273,
      "eval_runtime": 8.6274,
      "eval_samples_per_second": 14.025,
      "eval_steps_per_second": 1.855,
      "step": 560
    },
    {
      "epoch": 9.42,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1298026293516159,
      "eval_runtime": 8.3157,
      "eval_samples_per_second": 14.551,
      "eval_steps_per_second": 1.924,
      "step": 565
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.014311849139630795,
      "learning_rate": 1.0333333333333333e-05,
      "loss": 0.0058,
      "step": 570
    },
    {
      "epoch": 9.5,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13134212791919708,
      "eval_runtime": 9.1247,
      "eval_samples_per_second": 13.261,
      "eval_steps_per_second": 1.753,
      "step": 570
    },
    {
      "epoch": 9.58,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13212385773658752,
      "eval_runtime": 8.6468,
      "eval_samples_per_second": 13.994,
      "eval_steps_per_second": 1.85,
      "step": 575
    },
    {
      "epoch": 9.67,
      "grad_norm": 0.4894683361053467,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.012,
      "step": 580
    },
    {
      "epoch": 9.67,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1326775997877121,
      "eval_runtime": 9.2181,
      "eval_samples_per_second": 13.126,
      "eval_steps_per_second": 1.736,
      "step": 580
    },
    {
      "epoch": 9.75,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13264916837215424,
      "eval_runtime": 8.9299,
      "eval_samples_per_second": 13.55,
      "eval_steps_per_second": 1.792,
      "step": 585
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.015242637135088444,
      "learning_rate": 3.666666666666667e-06,
      "loss": 0.0081,
      "step": 590
    },
    {
      "epoch": 9.83,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13293592631816864,
      "eval_runtime": 9.2123,
      "eval_samples_per_second": 13.135,
      "eval_steps_per_second": 1.737,
      "step": 590
    },
    {
      "epoch": 9.92,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.13364511728286743,
      "eval_runtime": 8.9976,
      "eval_samples_per_second": 13.448,
      "eval_steps_per_second": 1.778,
      "step": 595
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.014774391427636147,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.0083,
      "step": 600
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9669421487603306,
      "eval_loss": 0.1337580680847168,
      "eval_runtime": 8.6721,
      "eval_samples_per_second": 13.953,
      "eval_steps_per_second": 1.845,
      "step": 600
    },
    {
      "epoch": 10.0,
      "step": 600,
      "total_flos": 7.39286832673751e+17,
      "train_loss": 0.06743512197087208,
      "train_runtime": 3546.8991,
      "train_samples_per_second": 2.69,
      "train_steps_per_second": 0.169
    }
  ],
  "logging_steps": 10,
  "max_steps": 600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 5,
  "total_flos": 7.39286832673751e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}