trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 253,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 1.8817,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.232421875,
      "eval_runtime": 77.7676,
      "eval_samples_per_second": 183.611,
      "eval_steps_per_second": 0.36,
      "step": 253
    },
    {
      "epoch": 1.0,
      "step": 253,
      "total_flos": 2542139602894848.0,
      "train_loss": 1.3124533189615242,
      "train_runtime": 2403.4192,
      "train_samples_per_second": 53.713,
      "train_steps_per_second": 0.105
    }
  ],
  "logging_steps": 500,
  "max_steps": 253,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 2542139602894848.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}