{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.2067400217056274,
      "learning_rate": 2.4e-05,
      "loss": 1.4814,
      "step": 4
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9487990140914917,
      "learning_rate": 1.2e-05,
      "loss": 1.3757,
      "step": 8
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.457656592130661,
      "learning_rate": 0.0,
      "loss": 1.3112,
      "step": 12
    }
  ],
  "logging_steps": 4,
  "max_steps": 12,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 857663889997824.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}