checkpoint-192/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.1818181818181817,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.02,
      "learning_rate": 6.60377358490566e-05,
      "loss": 0.7157,
      "step": 70
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.642105263157896e-05,
      "loss": 0.1645,
      "step": 140
    }
  ],
  "logging_steps": 70,
  "max_steps": 1056,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 4.50547197935616e+16,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}