{
  "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": 20.136756896972656,
      "learning_rate": 4.75e-05,
      "loss": 0.3889,
      "step": 122
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8045112781954887,
      "eval_f1": 0.7109554944646705,
      "eval_loss": 0.4199941158294678,
      "eval_precision": 0.8255285412262157,
      "eval_recall": 0.6866703036915802,
      "eval_runtime": 1.6394,
      "eval_samples_per_second": 243.375,
      "eval_steps_per_second": 30.498,
      "step": 122
    },
    {
      "epoch": 2.0,
      "grad_norm": 24.683944702148438,
      "learning_rate": 4.5e-05,
      "loss": 0.2335,
      "step": 244
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8739355018846853,
      "eval_loss": 0.3136064410209656,
      "eval_precision": 0.864426651415499,
      "eval_recall": 0.886252045826514,
      "eval_runtime": 1.6497,
      "eval_samples_per_second": 241.866,
      "eval_steps_per_second": 30.309,
      "step": 244
    },
    {
      "epoch": 3.0,
      "grad_norm": 66.46725463867188,
      "learning_rate": 4.25e-05,
      "loss": 0.1411,
      "step": 366
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8751002084335417,
      "eval_loss": 0.35689812898635864,
      "eval_precision": 0.8780701754385964,
      "eval_recall": 0.8722949627204946,
      "eval_runtime": 1.6606,
      "eval_samples_per_second": 240.275,
      "eval_steps_per_second": 30.11,
      "step": 366
    },
    {
      "epoch": 4.0,
      "grad_norm": 42.06414031982422,
      "learning_rate": 4e-05,
      "loss": 0.1078,
      "step": 488
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9147869674185464,
      "eval_f1": 0.8991765265473572,
      "eval_loss": 0.35370269417762756,
      "eval_precision": 0.8922773722627737,
      "eval_recall": 0.9072104018912529,
      "eval_runtime": 1.653,
      "eval_samples_per_second": 241.378,
      "eval_steps_per_second": 30.248,
      "step": 488
    },
    {
      "epoch": 5.0,
      "grad_norm": 88.54315185546875,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0822,
      "step": 610
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8796992481203008,
      "eval_f1": 0.8439374185136896,
      "eval_loss": 0.5069139003753662,
      "eval_precision": 0.8794955044955045,
      "eval_recall": 0.822376795781051,
      "eval_runtime": 1.6524,
      "eval_samples_per_second": 241.466,
      "eval_steps_per_second": 30.259,
      "step": 610
    },
    {
      "epoch": 6.0,
      "grad_norm": 114.8245849609375,
      "learning_rate": 3.5e-05,
      "loss": 0.0529,
      "step": 732
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9072681704260651,
      "eval_f1": 0.888964101175568,
      "eval_loss": 0.42624175548553467,
      "eval_precision": 0.8862007168458781,
      "eval_recall": 0.8918894344426259,
      "eval_runtime": 1.6561,
      "eval_samples_per_second": 240.934,
      "eval_steps_per_second": 30.192,
      "step": 732
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.022069375962018967,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.0365,
      "step": 854
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8769602202215754,
      "eval_loss": 0.5586097836494446,
      "eval_precision": 0.8742831541218639,
      "eval_recall": 0.8797963266048372,
      "eval_runtime": 1.6532,
      "eval_samples_per_second": 241.352,
      "eval_steps_per_second": 30.245,
      "step": 854
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.0406961552798748,
      "learning_rate": 3e-05,
      "loss": 0.033,
      "step": 976
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8947368421052632,
      "eval_f1": 0.8674628282189181,
      "eval_loss": 0.5012311935424805,
      "eval_precision": 0.8869858462356303,
      "eval_recall": 0.8530187306783051,
      "eval_runtime": 1.6551,
      "eval_samples_per_second": 241.075,
      "eval_steps_per_second": 30.21,
      "step": 976
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.6461573243141174,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.0248,
      "step": 1098
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8922305764411027,
      "eval_f1": 0.8631217838765008,
      "eval_loss": 0.583283007144928,
      "eval_precision": 0.8872804935927859,
      "eval_recall": 0.8462447717766868,
      "eval_runtime": 1.6572,
      "eval_samples_per_second": 240.772,
      "eval_steps_per_second": 30.172,
      "step": 1098
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.12847253680229187,
      "learning_rate": 2.5e-05,
      "loss": 0.0123,
      "step": 1220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8805765113084321,
      "eval_loss": 0.6610547304153442,
      "eval_precision": 0.8857796167247387,
      "eval_recall": 0.8758410620112748,
      "eval_runtime": 1.6505,
      "eval_samples_per_second": 241.744,
      "eval_steps_per_second": 30.294,
      "step": 1220
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.003805552376434207,
      "learning_rate": 2.25e-05,
      "loss": 0.0088,
      "step": 1342
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8947368421052632,
      "eval_f1": 0.8682132746146587,
      "eval_loss": 0.6935672760009766,
      "eval_precision": 0.884741537654159,
      "eval_recall": 0.8555191853064193,
      "eval_runtime": 1.6547,
      "eval_samples_per_second": 241.138,
      "eval_steps_per_second": 30.218,
      "step": 1342
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.0037182692904025316,
      "learning_rate": 2e-05,
      "loss": 0.0074,
      "step": 1464
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8805765113084321,
      "eval_loss": 0.6789939403533936,
      "eval_precision": 0.8857796167247387,
      "eval_recall": 0.8758410620112748,
      "eval_runtime": 1.6567,
      "eval_samples_per_second": 240.838,
      "eval_steps_per_second": 30.18,
      "step": 1464
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.0025616472121328115,
      "learning_rate": 1.75e-05,
      "loss": 0.0141,
      "step": 1586
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8731122745782431,
      "eval_loss": 0.6981470584869385,
      "eval_precision": 0.8829705994654449,
      "eval_recall": 0.864793598836152,
      "eval_runtime": 1.6639,
      "eval_samples_per_second": 239.794,
      "eval_steps_per_second": 30.049,
      "step": 1586
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.006673410069197416,
      "learning_rate": 1.5e-05,
      "loss": 0.0034,
      "step": 1708
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8972431077694235,
      "eval_f1": 0.8751002084335417,
      "eval_loss": 0.7144644856452942,
      "eval_precision": 0.8780701754385964,
      "eval_recall": 0.8722949627204946,
      "eval_runtime": 1.6531,
      "eval_samples_per_second": 241.366,
      "eval_steps_per_second": 30.246,
      "step": 1708
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.0030696168541908264,
      "learning_rate": 1.25e-05,
      "loss": 0.0059,
      "step": 1830
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.8758710801393728,
      "eval_loss": 0.7303631901741028,
      "eval_precision": 0.8870983228779925,
      "eval_recall": 0.8665666484815421,
      "eval_runtime": 1.6541,
      "eval_samples_per_second": 241.215,
      "eval_steps_per_second": 30.227,
      "step": 1830
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.0017388605047017336,
      "learning_rate": 1e-05,
      "loss": 0.0056,
      "step": 1952
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.899749373433584,
      "eval_f1": 0.879667048676036,
      "eval_loss": 0.7517656683921814,
      "eval_precision": 0.8778361344537815,
      "eval_recall": 0.8815693762502272,
      "eval_runtime": 1.6536,
      "eval_samples_per_second": 241.288,
      "eval_steps_per_second": 30.237,
      "step": 1952
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.002333118114620447,
      "learning_rate": 7.5e-06,
      "loss": 0.0039,
      "step": 2074
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8793019197207679,
      "eval_loss": 0.7390431761741638,
      "eval_precision": 0.8893184421534936,
      "eval_recall": 0.8708401527550463,
      "eval_runtime": 1.655,
      "eval_samples_per_second": 241.08,
      "eval_steps_per_second": 30.211,
      "step": 2074
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.0018157872837036848,
      "learning_rate": 5e-06,
      "loss": 0.004,
      "step": 2196
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8799463033398397,
      "eval_loss": 0.764133095741272,
      "eval_precision": 0.8874803397294746,
      "eval_recall": 0.8733406073831607,
      "eval_runtime": 1.6667,
      "eval_samples_per_second": 239.389,
      "eval_steps_per_second": 29.999,
      "step": 2196
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.0015570241957902908,
      "learning_rate": 2.5e-06,
      "loss": 0.007,
      "step": 2318
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8799463033398397,
      "eval_loss": 0.7847548723220825,
      "eval_precision": 0.8874803397294746,
      "eval_recall": 0.8733406073831607,
      "eval_runtime": 1.664,
      "eval_samples_per_second": 239.788,
      "eval_steps_per_second": 30.049,
      "step": 2318
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.002853752113878727,
      "learning_rate": 0.0,
      "loss": 0.0042,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9022556390977443,
      "eval_f1": 0.8799463033398397,
      "eval_loss": 0.790817379951477,
      "eval_precision": 0.8874803397294746,
      "eval_recall": 0.8733406073831607,
      "eval_runtime": 1.6678,
      "eval_samples_per_second": 239.236,
      "eval_steps_per_second": 29.979,
      "step": 2440
    },
    {
      "epoch": 20.0,
      "step": 2440,
      "total_flos": 7584162436176000.0,
      "train_loss": 0.0588726386183598,
      "train_runtime": 864.0501,
      "train_samples_per_second": 84.208,
      "train_steps_per_second": 2.824
    }
  ],
  "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
}