trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.028409090909090908,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "step": 10,
      "total_flos": 508182128640.0,
      "train_loss": 94.46571044921875,
      "train_runtime": 4.0219,
      "train_samples_per_second": 4.973,
      "train_steps_per_second": 2.486
    }
  ],
  "max_steps": 10,
  "num_train_epochs": 1,
  "total_flos": 508182128640.0,
  "trial_name": null,
  "trial_params": null
}