checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5773195876288657,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.58,
      "learning_rate": 7.0446735395189e-06,
      "loss": 0.3528,
      "step": 500
    }
  ],
  "logging_steps": 500,
  "max_steps": 582,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.0997907103744e+17,
  "trial_name": null,
  "trial_params": null
}