checkpoint-94/trainer_state.json

{
  "best_global_step": null,
  "best_metric": Infinity,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 94,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": NaN,
      "eval_runtime": 461.4953,
      "eval_samples_per_second": 0.581,
      "eval_steps_per_second": 0.581,
      "step": 47
    },
    {
      "epoch": 1.0642570281124497,
      "grad_norm": 0.13864651322364807,
      "learning_rate": 0.00014603174603174603,
      "loss": 0.10578728675842285,
      "step": 50
    },
    {
      "epoch": 2.0,
      "eval_loss": NaN,
      "eval_runtime": 462.8797,
      "eval_samples_per_second": 0.579,
      "eval_steps_per_second": 0.579,
      "step": 94
    }
  ],
  "logging_steps": 50,
  "max_steps": 141,
  "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": 3.318644627000525e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}