{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 63,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.48,
      "grad_norm": 78838.859375,
      "learning_rate": 0.00018782215733702286,
      "loss": 1.2119,
      "step": 10
    },
    {
      "epoch": 0.95,
      "grad_norm": 79446.625,
      "learning_rate": 0.00015425462638657595,
      "loss": 0.7581,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7133347392082214,
      "eval_runtime": 104.4914,
      "eval_samples_per_second": 1.062,
      "eval_steps_per_second": 0.134,
      "step": 21
    },
    {
      "epoch": 1.43,
      "grad_norm": 60001.390625,
      "learning_rate": 0.00010747300935864243,
      "loss": 0.6176,
      "step": 30
    },
    {
      "epoch": 1.9,
      "grad_norm": 59219.640625,
      "learning_rate": 5.887128968693887e-05,
      "loss": 0.612,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.680975079536438,
      "eval_runtime": 104.4154,
      "eval_samples_per_second": 1.063,
      "eval_steps_per_second": 0.134,
      "step": 42
    },
    {
      "epoch": 2.38,
      "grad_norm": 56114.09765625,
      "learning_rate": 2.0286749277707782e-05,
      "loss": 0.5092,
      "step": 50
    },
    {
      "epoch": 2.86,
      "grad_norm": 58096.5703125,
      "learning_rate": 1.1169173774871478e-06,
      "loss": 0.4944,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.6883889436721802,
      "eval_runtime": 104.4068,
      "eval_samples_per_second": 1.063,
      "eval_steps_per_second": 0.134,
      "step": 63
    },
    {
      "epoch": 3.0,
      "step": 63,
      "total_flos": 4.081496890291814e+16,
      "train_loss": 0.6921328268353901,
      "train_runtime": 3110.5155,
      "train_samples_per_second": 0.321,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 10,
  "max_steps": 63,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 10,
  "total_flos": 4.081496890291814e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}