{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 2440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.972439765930176,
      "learning_rate": 4.75e-05,
      "loss": 0.5634,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7192982456140351,
      "eval_f1": 0.638095238095238,
      "eval_loss": 0.5084609985351562,
      "eval_precision": 0.6523297491039426,
      "eval_recall": 0.6313875250045463,
      "eval_runtime": 1.7151,
      "eval_samples_per_second": 232.636,
      "eval_steps_per_second": 29.152,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.525497913360596,
      "learning_rate": 4.5e-05,
      "loss": 0.4815,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7468671679197995,
      "eval_f1": 0.7146872455662148,
      "eval_loss": 0.4543021321296692,
      "eval_precision": 0.7073059114891784,
      "eval_recall": 0.7334060738316057,
      "eval_runtime": 1.713,
      "eval_samples_per_second": 232.927,
      "eval_steps_per_second": 29.189,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.549623489379883,
      "learning_rate": 4.25e-05,
      "loss": 0.4003,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8270676691729323,
      "eval_f1": 0.7898027898027897,
      "eval_loss": 0.3836011290550232,
      "eval_precision": 0.7921052631578948,
      "eval_recall": 0.7876432078559739,
      "eval_runtime": 1.7117,
      "eval_samples_per_second": 233.096,
      "eval_steps_per_second": 29.21,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.769104957580566,
      "learning_rate": 4e-05,
      "loss": 0.3683,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8370927318295739,
      "eval_f1": 0.8029928975654221,
      "eval_loss": 0.35736599564552307,
      "eval_precision": 0.8037650785914463,
      "eval_recall": 0.8022367703218767,
      "eval_runtime": 1.7125,
      "eval_samples_per_second": 232.998,
      "eval_steps_per_second": 29.198,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.723803997039795,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.3396,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8596491228070176,
      "eval_f1": 0.8298403801632752,
      "eval_loss": 0.34212788939476013,
      "eval_precision": 0.8315523576240049,
      "eval_recall": 0.8281960356428442,
      "eval_runtime": 1.7123,
      "eval_samples_per_second": 233.026,
      "eval_steps_per_second": 29.201,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.710361480712891,
      "learning_rate": 3.5e-05,
      "loss": 0.3082,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8521303258145363,
      "eval_f1": 0.8285395893573786,
      "eval_loss": 0.3541903495788574,
      "eval_precision": 0.8185863137144467,
      "eval_recall": 0.8428805237315875,
      "eval_runtime": 1.7126,
      "eval_samples_per_second": 232.974,
      "eval_steps_per_second": 29.195,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.925419330596924,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.2908,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8571428571428571,
      "eval_f1": 0.8254579780661698,
      "eval_loss": 0.32880380749702454,
      "eval_precision": 0.8299216027874565,
      "eval_recall": 0.8214220767412257,
      "eval_runtime": 1.7119,
      "eval_samples_per_second": 233.076,
      "eval_steps_per_second": 29.208,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 6.063595294952393,
      "learning_rate": 3e-05,
      "loss": 0.2705,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8596491228070176,
      "eval_f1": 0.8347043853938569,
      "eval_loss": 0.332290917634964,
      "eval_precision": 0.8277993283927745,
      "eval_recall": 0.8431987634115294,
      "eval_runtime": 1.7111,
      "eval_samples_per_second": 233.18,
      "eval_steps_per_second": 29.22,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 4.576409339904785,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2608,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8671679197994987,
      "eval_f1": 0.8359744037230948,
      "eval_loss": 0.3119361996650696,
      "eval_precision": 0.8448835433371515,
      "eval_recall": 0.828514275322786,
      "eval_runtime": 1.7115,
      "eval_samples_per_second": 233.133,
      "eval_steps_per_second": 29.215,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 6.930353164672852,
      "learning_rate": 2.5e-05,
      "loss": 0.2489,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8671679197994987,
      "eval_f1": 0.8409485773595975,
      "eval_loss": 0.306035578250885,
      "eval_precision": 0.8385304659498208,
      "eval_recall": 0.8435170030914712,
      "eval_runtime": 1.7123,
      "eval_samples_per_second": 233.017,
      "eval_steps_per_second": 29.2,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 11.753581047058105,
      "learning_rate": 2.25e-05,
      "loss": 0.2452,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8822055137844611,
      "eval_f1": 0.8545433391506689,
      "eval_loss": 0.3053016662597656,
      "eval_precision": 0.8639270714012982,
      "eval_recall": 0.846653937079469,
      "eval_runtime": 1.7113,
      "eval_samples_per_second": 233.152,
      "eval_steps_per_second": 29.217,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 21.677701950073242,
      "learning_rate": 2e-05,
      "loss": 0.2357,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8822055137844611,
      "eval_f1": 0.851257564821498,
      "eval_loss": 0.30789536237716675,
      "eval_precision": 0.8714882943143812,
      "eval_recall": 0.8366521185670122,
      "eval_runtime": 1.7109,
      "eval_samples_per_second": 233.216,
      "eval_steps_per_second": 29.225,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 2.892643690109253,
      "learning_rate": 1.75e-05,
      "loss": 0.2206,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8847117794486216,
      "eval_f1": 0.8539996181748759,
      "eval_loss": 0.3078624904155731,
      "eval_precision": 0.8759595959595959,
      "eval_recall": 0.8384251682124022,
      "eval_runtime": 1.716,
      "eval_samples_per_second": 232.512,
      "eval_steps_per_second": 29.137,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 8.2120943069458,
      "learning_rate": 1.5e-05,
      "loss": 0.2272,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8721804511278195,
      "eval_f1": 0.8484099018899409,
      "eval_loss": 0.2965851128101349,
      "eval_precision": 0.8430645161290322,
      "eval_recall": 0.8545644662665939,
      "eval_runtime": 1.7126,
      "eval_samples_per_second": 232.98,
      "eval_steps_per_second": 29.196,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.116047978401184,
      "learning_rate": 1.25e-05,
      "loss": 0.2125,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8822055137844611,
      "eval_f1": 0.8568221901555235,
      "eval_loss": 0.2929779589176178,
      "eval_precision": 0.8596491228070176,
      "eval_recall": 0.8541553009638116,
      "eval_runtime": 1.7099,
      "eval_samples_per_second": 233.345,
      "eval_steps_per_second": 29.241,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 3.386876344680786,
      "learning_rate": 1e-05,
      "loss": 0.2094,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8548827059465357,
      "eval_loss": 0.2945634126663208,
      "eval_precision": 0.8548827059465357,
      "eval_recall": 0.8548827059465357,
      "eval_runtime": 1.7109,
      "eval_samples_per_second": 233.211,
      "eval_steps_per_second": 29.224,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 2.8860654830932617,
      "learning_rate": 7.5e-06,
      "loss": 0.2067,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8847117794486216,
      "eval_f1": 0.8602260265626904,
      "eval_loss": 0.29596519470214844,
      "eval_precision": 0.8620943049601959,
      "eval_recall": 0.8584288052373159,
      "eval_runtime": 1.7108,
      "eval_samples_per_second": 233.226,
      "eval_steps_per_second": 29.226,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 8.721236228942871,
      "learning_rate": 5e-06,
      "loss": 0.2058,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.8514869535493182,
      "eval_loss": 0.29655298590660095,
      "eval_precision": 0.8523821128305106,
      "eval_recall": 0.8506092016730314,
      "eval_runtime": 1.7113,
      "eval_samples_per_second": 233.15,
      "eval_steps_per_second": 29.217,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 4.918961524963379,
      "learning_rate": 2.5e-06,
      "loss": 0.2003,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8533986527862829,
      "eval_loss": 0.29609641432762146,
      "eval_precision": 0.8572003218020917,
      "eval_recall": 0.8498817966903074,
      "eval_runtime": 1.7119,
      "eval_samples_per_second": 233.075,
      "eval_steps_per_second": 29.207,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 4.333648681640625,
      "learning_rate": 0.0,
      "loss": 0.2019,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8533986527862829,
      "eval_loss": 0.2955795228481293,
      "eval_precision": 0.8572003218020917,
      "eval_recall": 0.8498817966903074,
      "eval_runtime": 1.7141,
      "eval_samples_per_second": 232.777,
      "eval_steps_per_second": 29.17,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7662265464912000.0,
      "train_loss": 0.28487560397288836,
      "train_runtime": 616.8461,
      "train_samples_per_second": 117.955,
      "train_steps_per_second": 3.956
    }
  ],
  "logging_steps": 500,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7662265464912000.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}