trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 26,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.1072778522968292,
      "epoch": 0.39215686274509803,
      "grad_norm": 2.03125,
      "learning_rate": 1.9775552389476865e-05,
      "loss": 1.156,
      "mean_token_accuracy": 0.7429225817322731,
      "num_tokens": 327680.0,
      "step": 10
    },
    {
      "entropy": 0.27947324328124523,
      "epoch": 0.7843137254901961,
      "grad_norm": 0.8359375,
      "learning_rate": 1.8039971303669407e-05,
      "loss": 0.2723,
      "mean_token_accuracy": 0.9285478919744492,
      "num_tokens": 655360.0,
      "step": 20
    }
  ],
  "logging_steps": 10,
  "max_steps": 78,
  "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": 3.544857646517453e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}