checkpoint-38/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 38,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 10.215806007385254,
      "learning_rate": 4.5e-05,
      "loss": 1.5086,
      "step": 10
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 7.984120845794678,
      "learning_rate": 4.318181818181819e-05,
      "loss": 1.3946,
      "step": 20
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 7.379543304443359,
      "learning_rate": 3.560606060606061e-05,
      "loss": 1.1182,
      "step": 30
    }
  ],
  "logging_steps": 10,
  "max_steps": 76,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 78409206454272.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}