trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "global_step": 780,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.41,
      "learning_rate": 1.076923076923077e-05,
      "loss": 95047.952,
      "step": 500
    },
    {
      "epoch": 10.0,
      "step": 780,
      "total_flos": 390360542976000.0,
      "train_loss": 60928.27408236968,
      "train_runtime": 290.1511,
      "train_samples_per_second": 85.817,
      "train_steps_per_second": 2.688
    }
  ],
  "max_steps": 780,
  "num_train_epochs": 10,
  "total_flos": 390360542976000.0,
  "trial_name": null,
  "trial_params": null
}