trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "grad_norm": 83.9439697265625,
      "learning_rate": 4.555555555555556e-05,
      "loss": 3.2962,
      "step": 50
    },
    {
      "epoch": 0.8,
      "grad_norm": 82.1087417602539,
      "learning_rate": 3.466666666666667e-05,
      "loss": 2.6118,
      "step": 100
    },
    {
      "epoch": 1.2,
      "grad_norm": 99.78679656982422,
      "learning_rate": 2.3555555555555556e-05,
      "loss": 2.4529,
      "step": 150
    },
    {
      "epoch": 1.6,
      "grad_norm": 94.11151885986328,
      "learning_rate": 1.2444444444444445e-05,
      "loss": 2.1187,
      "step": 200
    },
    {
      "epoch": 2.0,
      "grad_norm": 85.35782623291016,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 2.0534,
      "step": 250
    },
    {
      "epoch": 2.0,
      "step": 250,
      "total_flos": 9.4894217943552e+17,
      "train_loss": 2.5065918884277343,
      "train_runtime": 246.2795,
      "train_samples_per_second": 8.064,
      "train_steps_per_second": 1.015
    }
  ],
  "logging_steps": 50,
  "max_steps": 250,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.4894217943552e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}