{
  "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": 6.880457878112793,
      "learning_rate": 4.75e-05,
      "loss": 0.371,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8847117794486216,
      "eval_f1": 0.8556621579112929,
      "eval_loss": 0.2788735032081604,
      "eval_precision": 0.871654421411703,
      "eval_recall": 0.8434260774686306,
      "eval_runtime": 1.6426,
      "eval_samples_per_second": 242.907,
      "eval_steps_per_second": 30.439,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.694629192352295,
      "learning_rate": 4.5e-05,
      "loss": 0.214,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8947368421052632,
      "eval_f1": 0.8760282890453928,
      "eval_loss": 0.2702818810939789,
      "eval_precision": 0.8682260305697083,
      "eval_recall": 0.8855246408437898,
      "eval_runtime": 1.6739,
      "eval_samples_per_second": 238.362,
      "eval_steps_per_second": 29.87,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.0087978839874268,
      "learning_rate": 4.25e-05,
      "loss": 0.141,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.87468671679198,
      "eval_f1": 0.843111041207927,
      "eval_loss": 0.4446346163749695,
      "eval_precision": 0.8585673051692468,
      "eval_recall": 0.831332969630842,
      "eval_runtime": 1.6711,
      "eval_samples_per_second": 238.76,
      "eval_steps_per_second": 29.92,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 15.80390739440918,
      "learning_rate": 4e-05,
      "loss": 0.093,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8696741854636592,
      "eval_f1": 0.8368354828562441,
      "eval_loss": 0.5896394848823547,
      "eval_precision": 0.8520237470480189,
      "eval_recall": 0.8252864157119476,
      "eval_runtime": 1.6746,
      "eval_samples_per_second": 238.269,
      "eval_steps_per_second": 29.858,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.24515673518180847,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0469,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8563025210084034,
      "eval_loss": 0.6098523736000061,
      "eval_precision": 0.8529936381473334,
      "eval_recall": 0.8598836152027641,
      "eval_runtime": 1.6766,
      "eval_samples_per_second": 237.984,
      "eval_steps_per_second": 29.823,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.06188611686229706,
      "learning_rate": 3.5e-05,
      "loss": 0.0498,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.871530684525726,
      "eval_loss": 0.6609992980957031,
      "eval_precision": 0.9016447368421052,
      "eval_recall": 0.8515639207128569,
      "eval_runtime": 1.6777,
      "eval_samples_per_second": 237.819,
      "eval_steps_per_second": 29.802,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.2003283500671387,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.0257,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8532332818362393,
      "eval_loss": 0.6781011819839478,
      "eval_precision": 0.891747572815534,
      "eval_recall": 0.8301963993453355,
      "eval_runtime": 1.6733,
      "eval_samples_per_second": 238.446,
      "eval_steps_per_second": 29.88,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.8772869110107422,
      "learning_rate": 3e-05,
      "loss": 0.0267,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8523012380208119,
      "eval_loss": 0.8199848532676697,
      "eval_precision": 0.8950599239988307,
      "eval_recall": 0.8276959447172213,
      "eval_runtime": 1.6789,
      "eval_samples_per_second": 237.65,
      "eval_steps_per_second": 29.781,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.237112522125244,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.016,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.881931703852755,
      "eval_loss": 0.5965662002563477,
      "eval_precision": 0.8740012737378415,
      "eval_recall": 0.8915711947626841,
      "eval_runtime": 1.673,
      "eval_samples_per_second": 238.498,
      "eval_steps_per_second": 29.887,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.010850044898688793,
      "learning_rate": 2.5e-05,
      "loss": 0.0132,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8835263209107715,
      "eval_loss": 0.6436753273010254,
      "eval_precision": 0.8791501449961532,
      "eval_recall": 0.8883433351518457,
      "eval_runtime": 1.6736,
      "eval_samples_per_second": 238.404,
      "eval_steps_per_second": 29.875,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.008115105330944061,
      "learning_rate": 2.25e-05,
      "loss": 0.0161,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9072681704260651,
      "eval_f1": 0.8867007927797945,
      "eval_loss": 0.679655909538269,
      "eval_precision": 0.89198606271777,
      "eval_recall": 0.8818876159301692,
      "eval_runtime": 1.6719,
      "eval_samples_per_second": 238.648,
      "eval_steps_per_second": 29.906,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.003001323202624917,
      "learning_rate": 2e-05,
      "loss": 0.0091,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9097744360902256,
      "eval_f1": 0.8882839721254355,
      "eval_loss": 0.6953954696655273,
      "eval_precision": 0.8998687748047625,
      "eval_recall": 0.8786597563193308,
      "eval_runtime": 1.6653,
      "eval_samples_per_second": 239.59,
      "eval_steps_per_second": 30.024,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.0026766008231788874,
      "learning_rate": 1.75e-05,
      "loss": 0.0101,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9122807017543859,
      "eval_f1": 0.8954723392788977,
      "eval_loss": 0.6750524044036865,
      "eval_precision": 0.8909569746108776,
      "eval_recall": 0.9004364429896345,
      "eval_runtime": 1.6681,
      "eval_samples_per_second": 239.196,
      "eval_steps_per_second": 29.974,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.0015771281905472279,
      "learning_rate": 1.5e-05,
      "loss": 0.0025,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8779692119482108,
      "eval_loss": 0.7316891551017761,
      "eval_precision": 0.8934322033898305,
      "eval_recall": 0.865839243498818,
      "eval_runtime": 1.6607,
      "eval_samples_per_second": 240.257,
      "eval_steps_per_second": 30.107,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.003234785981476307,
      "learning_rate": 1.25e-05,
      "loss": 0.0088,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8897243107769424,
      "eval_f1": 0.8669758137843244,
      "eval_loss": 0.6788834929466248,
      "eval_precision": 0.8669758137843244,
      "eval_recall": 0.8669758137843244,
      "eval_runtime": 1.6627,
      "eval_samples_per_second": 239.966,
      "eval_steps_per_second": 30.071,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.0018662125803530216,
      "learning_rate": 1e-05,
      "loss": 0.0017,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8897243107769424,
      "eval_f1": 0.8676337535436396,
      "eval_loss": 0.7504969239234924,
      "eval_precision": 0.8658613445378152,
      "eval_recall": 0.8694762684124386,
      "eval_runtime": 1.6676,
      "eval_samples_per_second": 239.272,
      "eval_steps_per_second": 29.984,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.0024691985454410315,
      "learning_rate": 7.5e-06,
      "loss": 0.0017,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8897243107769424,
      "eval_f1": 0.8676337535436396,
      "eval_loss": 0.7755724787712097,
      "eval_precision": 0.8658613445378152,
      "eval_recall": 0.8694762684124386,
      "eval_runtime": 1.6752,
      "eval_samples_per_second": 238.182,
      "eval_steps_per_second": 29.847,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.0015471646329388022,
      "learning_rate": 5e-06,
      "loss": 0.0011,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8715803025426456,
      "eval_loss": 0.8041408061981201,
      "eval_precision": 0.8673433153814287,
      "eval_recall": 0.8762502273140571,
      "eval_runtime": 1.6768,
      "eval_samples_per_second": 237.95,
      "eval_steps_per_second": 29.818,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.0015901036094874144,
      "learning_rate": 2.5e-06,
      "loss": 0.0017,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8703223612108386,
      "eval_loss": 0.8064257502555847,
      "eval_precision": 0.8694131129742446,
      "eval_recall": 0.8712493180578287,
      "eval_runtime": 1.6795,
      "eval_samples_per_second": 237.568,
      "eval_steps_per_second": 29.77,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0011909452732652426,
      "learning_rate": 0.0,
      "loss": 0.0008,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8703223612108386,
      "eval_loss": 0.8053188323974609,
      "eval_precision": 0.8694131129742446,
      "eval_recall": 0.8712493180578287,
      "eval_runtime": 1.6806,
      "eval_samples_per_second": 237.419,
      "eval_steps_per_second": 29.752,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7598755382040000.0,
      "train_loss": 0.05254678238855034,
      "train_runtime": 869.5662,
      "train_samples_per_second": 83.835,
      "train_steps_per_second": 2.806
    }
  ],
  "logging_steps": 500,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7598755382040000.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}