{
  "best_metric": 0.4286545217037201,
  "best_model_checkpoint": "./vit-augmentation/checkpoint-1926",
  "epoch": 16.0,
  "eval_steps": 500,
  "global_step": 5136,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 6.7570695877075195,
      "learning_rate": 4.1428571428571437e-05,
      "loss": 0.9124,
      "step": 321
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7805131761442441,
      "eval_f1": 0.768284202243805,
      "eval_loss": 0.6024736166000366,
      "eval_precision": 0.778781843539936,
      "eval_recall": 0.7805131761442441,
      "eval_runtime": 24.6229,
      "eval_samples_per_second": 117.127,
      "eval_steps_per_second": 14.661,
      "step": 321
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.467956066131592,
      "learning_rate": 8.311688311688312e-05,
      "loss": 0.5876,
      "step": 642
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7864077669902912,
      "eval_f1": 0.7820269177484748,
      "eval_loss": 0.5819488167762756,
      "eval_precision": 0.7990366623686466,
      "eval_recall": 0.7864077669902912,
      "eval_runtime": 24.6924,
      "eval_samples_per_second": 116.797,
      "eval_steps_per_second": 14.62,
      "step": 642
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.2375881671905518,
      "learning_rate": 9.985334621908699e-05,
      "loss": 0.5415,
      "step": 963
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8040915395284327,
      "eval_f1": 0.7864667726550387,
      "eval_loss": 0.6148691773414612,
      "eval_precision": 0.7943404456778191,
      "eval_recall": 0.8040915395284327,
      "eval_runtime": 23.4113,
      "eval_samples_per_second": 123.188,
      "eval_steps_per_second": 15.42,
      "step": 963
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.80470085144043,
      "learning_rate": 9.894936461151184e-05,
      "loss": 0.4815,
      "step": 1284
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8294036061026352,
      "eval_f1": 0.8114752843394228,
      "eval_loss": 0.4653749465942383,
      "eval_precision": 0.8259066316761086,
      "eval_recall": 0.8294036061026352,
      "eval_runtime": 22.9294,
      "eval_samples_per_second": 125.778,
      "eval_steps_per_second": 15.744,
      "step": 1284
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.0819091796875,
      "learning_rate": 9.723506398349735e-05,
      "loss": 0.4263,
      "step": 1605
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8259361997226075,
      "eval_f1": 0.8022540151903795,
      "eval_loss": 0.5480949282646179,
      "eval_precision": 0.8315275661759924,
      "eval_recall": 0.8259361997226075,
      "eval_runtime": 22.776,
      "eval_samples_per_second": 126.625,
      "eval_steps_per_second": 15.85,
      "step": 1605
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.5868643522262573,
      "learning_rate": 9.473882326123909e-05,
      "loss": 0.3515,
      "step": 1926
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8592233009708737,
      "eval_f1": 0.857445188937737,
      "eval_loss": 0.4286545217037201,
      "eval_precision": 0.8579724186747698,
      "eval_recall": 0.8592233009708737,
      "eval_runtime": 22.7677,
      "eval_samples_per_second": 126.671,
      "eval_steps_per_second": 15.856,
      "step": 1926
    },
    {
      "epoch": 7.0,
      "grad_norm": 7.5292840003967285,
      "learning_rate": 9.15019657867844e-05,
      "loss": 0.3144,
      "step": 2247
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8363384188626907,
      "eval_f1": 0.8269755362382184,
      "eval_loss": 0.5004583597183228,
      "eval_precision": 0.8320109172887173,
      "eval_recall": 0.8363384188626907,
      "eval_runtime": 22.8371,
      "eval_samples_per_second": 126.286,
      "eval_steps_per_second": 15.808,
      "step": 2247
    },
    {
      "epoch": 8.0,
      "grad_norm": 9.491310119628906,
      "learning_rate": 8.757807524194037e-05,
      "loss": 0.2736,
      "step": 2568
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8294036061026352,
      "eval_f1": 0.8301623012381248,
      "eval_loss": 0.5306037664413452,
      "eval_precision": 0.8447810859364984,
      "eval_recall": 0.8294036061026352,
      "eval_runtime": 22.6737,
      "eval_samples_per_second": 127.196,
      "eval_steps_per_second": 15.922,
      "step": 2568
    },
    {
      "epoch": 9.0,
      "grad_norm": 5.960662841796875,
      "learning_rate": 8.303210861170248e-05,
      "loss": 0.2519,
      "step": 2889
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8578363384188626,
      "eval_f1": 0.8534043679644312,
      "eval_loss": 0.4732687175273895,
      "eval_precision": 0.8533865418942713,
      "eval_recall": 0.8578363384188626,
      "eval_runtime": 22.5796,
      "eval_samples_per_second": 127.726,
      "eval_steps_per_second": 15.988,
      "step": 2889
    },
    {
      "epoch": 10.0,
      "grad_norm": 8.973663330078125,
      "learning_rate": 7.793932087141109e-05,
      "loss": 0.2227,
      "step": 3210
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8585298196948682,
      "eval_f1": 0.851216238380984,
      "eval_loss": 0.4905049800872803,
      "eval_precision": 0.8519657875781843,
      "eval_recall": 0.8585298196948682,
      "eval_runtime": 24.7137,
      "eval_samples_per_second": 116.696,
      "eval_steps_per_second": 14.607,
      "step": 3210
    },
    {
      "epoch": 11.0,
      "grad_norm": 3.6425747871398926,
      "learning_rate": 7.23840191986112e-05,
      "loss": 0.1724,
      "step": 3531
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8654646324549237,
      "eval_f1": 0.862832998223053,
      "eval_loss": 0.5050108432769775,
      "eval_precision": 0.8670507633275476,
      "eval_recall": 0.8654646324549237,
      "eval_runtime": 24.5636,
      "eval_samples_per_second": 117.409,
      "eval_steps_per_second": 14.697,
      "step": 3531
    },
    {
      "epoch": 12.0,
      "grad_norm": 14.814494132995605,
      "learning_rate": 6.64581673326787e-05,
      "loss": 0.1596,
      "step": 3852
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8685852981969486,
      "eval_f1": 0.8631462417126987,
      "eval_loss": 0.526269257068634,
      "eval_precision": 0.8656616085283582,
      "eval_recall": 0.8685852981969486,
      "eval_runtime": 24.7987,
      "eval_samples_per_second": 116.296,
      "eval_steps_per_second": 14.557,
      "step": 3852
    },
    {
      "epoch": 13.0,
      "grad_norm": 1.3064632415771484,
      "learning_rate": 6.027949045818934e-05,
      "loss": 0.1397,
      "step": 4173
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8533287101248266,
      "eval_f1": 0.848769203208603,
      "eval_loss": 0.7043444514274597,
      "eval_precision": 0.8702783792779959,
      "eval_recall": 0.8533287101248266,
      "eval_runtime": 24.4467,
      "eval_samples_per_second": 117.971,
      "eval_steps_per_second": 14.767,
      "step": 4173
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.027323821559548378,
      "learning_rate": 5.391170860718704e-05,
      "loss": 0.1298,
      "step": 4494
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8678918169209431,
      "eval_f1": 0.8631936421240504,
      "eval_loss": 0.6274930834770203,
      "eval_precision": 0.8734415776541141,
      "eval_recall": 0.8678918169209431,
      "eval_runtime": 24.4385,
      "eval_samples_per_second": 118.01,
      "eval_steps_per_second": 14.772,
      "step": 4494
    },
    {
      "epoch": 15.0,
      "grad_norm": 18.411346435546875,
      "learning_rate": 4.7479171433782145e-05,
      "loss": 0.1029,
      "step": 4815
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8807212205270458,
      "eval_f1": 0.8772158509937853,
      "eval_loss": 0.5564317107200623,
      "eval_precision": 0.8775761761640355,
      "eval_recall": 0.8807212205270458,
      "eval_runtime": 24.4883,
      "eval_samples_per_second": 117.771,
      "eval_steps_per_second": 14.742,
      "step": 4815
    },
    {
      "epoch": 16.0,
      "grad_norm": 17.630950927734375,
      "learning_rate": 4.1088364635238816e-05,
      "loss": 0.0893,
      "step": 5136
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.880374479889043,
      "eval_f1": 0.8789241435580383,
      "eval_loss": 0.5668464303016663,
      "eval_precision": 0.8823340407710801,
      "eval_recall": 0.880374479889043,
      "eval_runtime": 24.4769,
      "eval_samples_per_second": 117.825,
      "eval_steps_per_second": 14.749,
      "step": 5136
    },
    {
      "epoch": 16.0,
      "step": 5136,
      "total_flos": 6.35834769042545e+18,
      "train_loss": 0.3223209912160475,
      "train_runtime": 1869.7645,
      "train_samples_per_second": 274.259,
      "train_steps_per_second": 17.168
    }
  ],
  "logging_steps": 500,
  "max_steps": 32100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 6.35834769042545e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}