trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.32051282051282054,
      "grad_norm": 2.562042009317436,
      "learning_rate": 7.607142857142856e-07,
      "loss": 0.0514,
      "step": 100
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 1.8491213516223084,
      "learning_rate": 4.0357142857142854e-07,
      "loss": 0.0291,
      "step": 200
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 1.4051631542245315,
      "learning_rate": 4.642857142857143e-08,
      "loss": 0.0276,
      "step": 300
    }
  ],
  "logging_steps": 100,
  "max_steps": 312,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}