{
  "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": 4.480478763580322,
      "learning_rate": 4.75e-05,
      "loss": 0.5517,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7167919799498746,
      "eval_f1": 0.6423852090389206,
      "eval_loss": 0.5130985379219055,
      "eval_precision": 0.6513043478260869,
      "eval_recall": 0.6371158392434988,
      "eval_runtime": 1.7122,
      "eval_samples_per_second": 233.038,
      "eval_steps_per_second": 29.203,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.928369998931885,
      "learning_rate": 4.5e-05,
      "loss": 0.4833,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7518796992481203,
      "eval_f1": 0.7158957687522027,
      "eval_loss": 0.4657236337661743,
      "eval_precision": 0.7087837837837838,
      "eval_recall": 0.7294508092380433,
      "eval_runtime": 1.7104,
      "eval_samples_per_second": 233.285,
      "eval_steps_per_second": 29.234,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.609431266784668,
      "learning_rate": 4.25e-05,
      "loss": 0.4318,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8120300751879699,
      "eval_f1": 0.7781114447781114,
      "eval_loss": 0.40558183193206787,
      "eval_precision": 0.7728937728937728,
      "eval_recall": 0.784506273867976,
      "eval_runtime": 1.7095,
      "eval_samples_per_second": 233.398,
      "eval_steps_per_second": 29.248,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.862994194030762,
      "learning_rate": 4e-05,
      "loss": 0.3905,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8421052631578947,
      "eval_f1": 0.8100071803786705,
      "eval_loss": 0.3810623586177826,
      "eval_precision": 0.8092466373122624,
      "eval_recall": 0.8107837788688852,
      "eval_runtime": 1.71,
      "eval_samples_per_second": 233.338,
      "eval_steps_per_second": 29.24,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.350956916809082,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.3626,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.849624060150376,
      "eval_f1": 0.8186033824331697,
      "eval_loss": 0.36519819498062134,
      "eval_precision": 0.8186033824331697,
      "eval_recall": 0.8186033824331697,
      "eval_runtime": 1.7101,
      "eval_samples_per_second": 233.324,
      "eval_steps_per_second": 29.239,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.2842788696289062,
      "learning_rate": 3.5e-05,
      "loss": 0.3331,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8546365914786967,
      "eval_f1": 0.8325179481241316,
      "eval_loss": 0.3646066188812256,
      "eval_precision": 0.8214046915919095,
      "eval_recall": 0.849654482633206,
      "eval_runtime": 1.7082,
      "eval_samples_per_second": 233.581,
      "eval_steps_per_second": 29.271,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.8412091732025146,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.3134,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8671679197994987,
      "eval_f1": 0.8385441718775052,
      "eval_loss": 0.34403207898139954,
      "eval_precision": 0.8412280701754387,
      "eval_recall": 0.8360156392071285,
      "eval_runtime": 1.7096,
      "eval_samples_per_second": 233.394,
      "eval_steps_per_second": 29.247,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 5.936851501464844,
      "learning_rate": 3e-05,
      "loss": 0.2927,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8646616541353384,
      "eval_f1": 0.8375505157126486,
      "eval_loss": 0.3412378430366516,
      "eval_precision": 0.8359243697478991,
      "eval_recall": 0.8392434988179669,
      "eval_runtime": 1.7101,
      "eval_samples_per_second": 233.324,
      "eval_steps_per_second": 29.239,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 10.900829315185547,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.2833,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8646616541353384,
      "eval_f1": 0.8383403361344538,
      "eval_loss": 0.3352757394313812,
      "eval_precision": 0.8352012604792199,
      "eval_recall": 0.8417439534460811,
      "eval_runtime": 1.709,
      "eval_samples_per_second": 233.472,
      "eval_steps_per_second": 29.257,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 5.930834770202637,
      "learning_rate": 2.5e-05,
      "loss": 0.2672,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8671679197994987,
      "eval_f1": 0.8431987543098654,
      "eval_loss": 0.3296276330947876,
      "eval_precision": 0.8366910866910866,
      "eval_recall": 0.8510183669758138,
      "eval_runtime": 1.7081,
      "eval_samples_per_second": 233.589,
      "eval_steps_per_second": 29.272,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 8.893163681030273,
      "learning_rate": 2.25e-05,
      "loss": 0.2641,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.8483536940081443,
      "eval_loss": 0.3269977569580078,
      "eval_precision": 0.8575792287132493,
      "eval_recall": 0.8406073831605747,
      "eval_runtime": 1.7082,
      "eval_samples_per_second": 233.574,
      "eval_steps_per_second": 29.27,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 21.00832748413086,
      "learning_rate": 2e-05,
      "loss": 0.2549,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8696741854636592,
      "eval_f1": 0.8349560901107294,
      "eval_loss": 0.33516642451286316,
      "eval_precision": 0.8558080808080808,
      "eval_recall": 0.8202855064557192,
      "eval_runtime": 1.708,
      "eval_samples_per_second": 233.601,
      "eval_steps_per_second": 29.273,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 2.7979023456573486,
      "learning_rate": 1.75e-05,
      "loss": 0.2534,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8696741854636592,
      "eval_f1": 0.8329898898834439,
      "eval_loss": 0.34021762013435364,
      "eval_precision": 0.8601973684210527,
      "eval_recall": 0.8152845971994909,
      "eval_runtime": 1.7123,
      "eval_samples_per_second": 233.024,
      "eval_steps_per_second": 29.201,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 10.50624942779541,
      "learning_rate": 1.5e-05,
      "loss": 0.2389,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8822055137844611,
      "eval_f1": 0.858259325044405,
      "eval_loss": 0.3207694888114929,
      "eval_precision": 0.8573798178418481,
      "eval_recall": 0.8591562102200401,
      "eval_runtime": 1.7089,
      "eval_samples_per_second": 233.48,
      "eval_steps_per_second": 29.258,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 3.9172396659851074,
      "learning_rate": 1.25e-05,
      "loss": 0.2203,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.87468671679198,
      "eval_f1": 0.8422176526415692,
      "eval_loss": 0.3279098570346832,
      "eval_precision": 0.8604724566416373,
      "eval_recall": 0.8288325150027278,
      "eval_runtime": 1.7092,
      "eval_samples_per_second": 233.448,
      "eval_steps_per_second": 29.254,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 3.250375986099243,
      "learning_rate": 1e-05,
      "loss": 0.2298,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.87468671679198,
      "eval_f1": 0.8448388501742161,
      "eval_loss": 0.3175312280654907,
      "eval_precision": 0.8551721930610677,
      "eval_recall": 0.8363338788870704,
      "eval_runtime": 1.7091,
      "eval_samples_per_second": 233.452,
      "eval_steps_per_second": 29.255,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.9103662967681885,
      "learning_rate": 7.5e-06,
      "loss": 0.2227,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.87468671679198,
      "eval_f1": 0.843111041207927,
      "eval_loss": 0.32178837060928345,
      "eval_precision": 0.8585673051692468,
      "eval_recall": 0.831332969630842,
      "eval_runtime": 1.7084,
      "eval_samples_per_second": 233.545,
      "eval_steps_per_second": 29.266,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 7.885207176208496,
      "learning_rate": 5e-06,
      "loss": 0.2225,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.8514869535493182,
      "eval_loss": 0.3177903890609741,
      "eval_precision": 0.8523821128305106,
      "eval_recall": 0.8506092016730314,
      "eval_runtime": 1.709,
      "eval_samples_per_second": 233.472,
      "eval_steps_per_second": 29.257,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 4.890872478485107,
      "learning_rate": 2.5e-06,
      "loss": 0.2192,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.846679266293906,
      "eval_loss": 0.3198798894882202,
      "eval_precision": 0.8609191655801824,
      "eval_recall": 0.8356064739043463,
      "eval_runtime": 1.7082,
      "eval_samples_per_second": 233.574,
      "eval_steps_per_second": 29.27,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 4.476297378540039,
      "learning_rate": 0.0,
      "loss": 0.2229,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.846679266293906,
      "eval_loss": 0.3205019235610962,
      "eval_precision": 0.8609191655801824,
      "eval_recall": 0.8356064739043463,
      "eval_runtime": 1.7082,
      "eval_samples_per_second": 233.579,
      "eval_steps_per_second": 29.271,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7662265464912000.0,
      "train_loss": 0.3029043838626049,
      "train_runtime": 615.7228,
      "train_samples_per_second": 118.17,
      "train_steps_per_second": 3.963
    }
  ],
  "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
}