checkpoint-129/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 129,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 5.726939678192139,
      "learning_rate": 0.00018604651162790697,
      "loss": 2.4693,
      "step": 50
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 2.5614702701568604,
      "learning_rate": 6.976744186046511e-05,
      "loss": 0.8933,
      "step": 100
    }
  ],
  "logging_steps": 50,
  "max_steps": 129,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 308221779640320.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}