trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 63,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 0.24600112438201904,
      "learning_rate": 0.00029986484493070223,
      "loss": 0.7743,
      "step": 20
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 0.3101575970649719,
      "learning_rate": 0.00024423299959434636,
      "loss": 0.5663,
      "step": 40
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.28579050302505493,
      "learning_rate": 0.00012464987695172264,
      "loss": 0.4193,
      "step": 60
    }
  ],
  "logging_steps": 20,
  "max_steps": 93,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 5.130438997144781e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}