trainer_state.json

{
  "best_metric": 67.28861326226105,
  "best_model_checkpoint": "./model/checkpoint-5285",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5285,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.05830267071723938,
      "learning_rate": 1.5e-05,
      "loss": 0.132,
      "step": 5285
    },
    {
      "epoch": 1.0,
      "eval_bleu": 67.28861326226105,
      "eval_loss": 0.02794693037867546,
      "eval_runtime": 863.2482,
      "eval_samples_per_second": 10.884,
      "eval_steps_per_second": 0.681,
      "step": 5285
    }
  ],
  "logging_steps": 500,
  "max_steps": 10570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.149104339419136e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}