{
  "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": 21.05661964416504,
      "learning_rate": 4.75e-05,
      "loss": 0.411,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8721804511278195,
      "eval_f1": 0.8446368446368446,
      "eval_loss": 0.275076687335968,
      "eval_precision": 0.8473684210526315,
      "eval_recall": 0.8420621931260229,
      "eval_runtime": 4.712,
      "eval_samples_per_second": 84.677,
      "eval_steps_per_second": 10.611,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 21.74665069580078,
      "learning_rate": 4.5e-05,
      "loss": 0.2264,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8718936621074937,
      "eval_loss": 0.3036659359931946,
      "eval_precision": 0.8573529411764707,
      "eval_recall": 0.897708674304419,
      "eval_runtime": 4.9592,
      "eval_samples_per_second": 80.457,
      "eval_steps_per_second": 10.082,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.0689539909362793,
      "learning_rate": 4.25e-05,
      "loss": 0.1467,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8771929824561403,
      "eval_f1": 0.8582079268956014,
      "eval_loss": 0.3442274332046509,
      "eval_precision": 0.8464828897338403,
      "eval_recall": 0.8756137479541735,
      "eval_runtime": 4.9511,
      "eval_samples_per_second": 80.588,
      "eval_steps_per_second": 10.099,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.8201183080673218,
      "learning_rate": 4e-05,
      "loss": 0.0961,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8778322106552358,
      "eval_loss": 0.3736657500267029,
      "eval_precision": 0.8818924438393465,
      "eval_recall": 0.8740680123658847,
      "eval_runtime": 4.9528,
      "eval_samples_per_second": 80.56,
      "eval_steps_per_second": 10.095,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0896231159567833,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0726,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8771929824561404,
      "eval_loss": 0.43064403533935547,
      "eval_precision": 0.8834928678678678,
      "eval_recall": 0.8715675577377705,
      "eval_runtime": 4.9624,
      "eval_samples_per_second": 80.405,
      "eval_steps_per_second": 10.076,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 53.643978118896484,
      "learning_rate": 3.5e-05,
      "loss": 0.0514,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8847117794486216,
      "eval_f1": 0.8677208256457565,
      "eval_loss": 0.6448621153831482,
      "eval_precision": 0.8546209186496956,
      "eval_recall": 0.8884342607746863,
      "eval_runtime": 4.9531,
      "eval_samples_per_second": 80.556,
      "eval_steps_per_second": 10.095,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03521590679883957,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.0532,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8763538792940554,
      "eval_loss": 0.5595228672027588,
      "eval_precision": 0.8754297605404427,
      "eval_recall": 0.877295871976723,
      "eval_runtime": 4.9653,
      "eval_samples_per_second": 80.357,
      "eval_steps_per_second": 10.07,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.004897149745374918,
      "learning_rate": 3e-05,
      "loss": 0.0274,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8872180451127819,
      "eval_f1": 0.8614765038536611,
      "eval_loss": 0.6727907657623291,
      "eval_precision": 0.8686536646744258,
      "eval_recall": 0.8552009456264775,
      "eval_runtime": 4.9483,
      "eval_samples_per_second": 80.635,
      "eval_steps_per_second": 10.105,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.02297130785882473,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.0186,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9072681704260651,
      "eval_f1": 0.8848664457009163,
      "eval_loss": 0.6217536926269531,
      "eval_precision": 0.8977236138837015,
      "eval_recall": 0.8743862520458265,
      "eval_runtime": 4.9609,
      "eval_samples_per_second": 80.43,
      "eval_steps_per_second": 10.079,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.018355082720518112,
      "learning_rate": 2.5e-05,
      "loss": 0.0121,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8669226294357184,
      "eval_loss": 0.6576113104820251,
      "eval_precision": 0.8766227567773959,
      "eval_recall": 0.8587470449172576,
      "eval_runtime": 4.9593,
      "eval_samples_per_second": 80.455,
      "eval_steps_per_second": 10.082,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.0026700079906731844,
      "learning_rate": 2.25e-05,
      "loss": 0.0244,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8694882125334078,
      "eval_loss": 0.7506579160690308,
      "eval_precision": 0.8940436639772188,
      "eval_recall": 0.8522913256955811,
      "eval_runtime": 4.9497,
      "eval_samples_per_second": 80.611,
      "eval_steps_per_second": 10.102,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.002809323836117983,
      "learning_rate": 2e-05,
      "loss": 0.0062,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8724195749658803,
      "eval_loss": 0.6859227418899536,
      "eval_precision": 0.8848766823362741,
      "eval_recall": 0.8622931442080378,
      "eval_runtime": 4.9227,
      "eval_samples_per_second": 81.053,
      "eval_steps_per_second": 10.157,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.004050145391374826,
      "learning_rate": 1.75e-05,
      "loss": 0.0099,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9072681704260651,
      "eval_f1": 0.8872855539522206,
      "eval_loss": 0.6514401435852051,
      "eval_precision": 0.8903508771929824,
      "eval_recall": 0.8843880705582834,
      "eval_runtime": 4.9442,
      "eval_samples_per_second": 80.7,
      "eval_steps_per_second": 10.113,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.0045247310772538185,
      "learning_rate": 1.5e-05,
      "loss": 0.0087,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8765393898137261,
      "eval_loss": 0.7604307532310486,
      "eval_precision": 0.8852261942423283,
      "eval_recall": 0.8690671031096563,
      "eval_runtime": 4.9491,
      "eval_samples_per_second": 80.62,
      "eval_steps_per_second": 10.103,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.002084016567096114,
      "learning_rate": 1.25e-05,
      "loss": 0.0056,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8799463033398397,
      "eval_loss": 0.7281515598297119,
      "eval_precision": 0.8874803397294746,
      "eval_recall": 0.8733406073831607,
      "eval_runtime": 4.9475,
      "eval_samples_per_second": 80.646,
      "eval_steps_per_second": 10.106,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.0016144708497449756,
      "learning_rate": 1e-05,
      "loss": 0.0063,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.9122807017543859,
      "eval_f1": 0.89337822671156,
      "eval_loss": 0.6987277269363403,
      "eval_precision": 0.8964912280701754,
      "eval_recall": 0.8904346244771777,
      "eval_runtime": 4.9546,
      "eval_samples_per_second": 80.531,
      "eval_steps_per_second": 10.092,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.0015528218355029821,
      "learning_rate": 7.5e-06,
      "loss": 0.0071,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9047619047619048,
      "eval_f1": 0.8833333333333333,
      "eval_loss": 0.7402216792106628,
      "eval_precision": 0.8897334834834836,
      "eval_recall": 0.8776141116566649,
      "eval_runtime": 4.977,
      "eval_samples_per_second": 80.17,
      "eval_steps_per_second": 10.046,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.0025696582160890102,
      "learning_rate": 5e-06,
      "loss": 0.0023,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8690075356742023,
      "eval_loss": 0.7846499085426331,
      "eval_precision": 0.8719298245614036,
      "eval_recall": 0.8662484088016003,
      "eval_runtime": 4.9435,
      "eval_samples_per_second": 80.712,
      "eval_steps_per_second": 10.114,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.0012161381309852004,
      "learning_rate": 2.5e-06,
      "loss": 0.0043,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8690075356742023,
      "eval_loss": 0.7948206067085266,
      "eval_precision": 0.8719298245614036,
      "eval_recall": 0.8662484088016003,
      "eval_runtime": 4.9539,
      "eval_samples_per_second": 80.542,
      "eval_steps_per_second": 10.093,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0017919199308380485,
      "learning_rate": 0.0,
      "loss": 0.0021,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8744522298370696,
      "eval_loss": 0.7891051769256592,
      "eval_precision": 0.8795731707317074,
      "eval_recall": 0.8697945080923805,
      "eval_runtime": 4.9391,
      "eval_samples_per_second": 80.783,
      "eval_steps_per_second": 10.123,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7584162436176000.0,
      "train_loss": 0.05962105130807298,
      "train_runtime": 2714.936,
      "train_samples_per_second": 26.8,
      "train_steps_per_second": 0.899
    }
  ],
  "logging_steps": 500,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7584162436176000.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}