trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9980806142034548,
  "global_step": 520,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.92,
      "learning_rate": 2.211538461538462e-06,
      "loss": 0.1134,
      "step": 500
    },
    {
      "epoch": 2.0,
      "step": 520,
      "total_flos": 751961665666944.0,
      "train_loss": 0.11060212357686117,
      "train_runtime": 43.5566,
      "train_samples_per_second": 382.215,
      "train_steps_per_second": 11.938
    }
  ],
  "max_steps": 520,
  "num_train_epochs": 2,
  "total_flos": 751961665666944.0,
  "trial_name": null,
  "trial_params": null
}