checkpoint-12838/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "global_step": 12838,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 8.889735156655704e-06,
      "loss": 7.3112,
      "step": 6419
    },
    {
      "epoch": 1.0,
      "eval_loss": 6.181427955627441,
      "eval_runtime": 64.709,
      "eval_samples_per_second": 198.396,
      "eval_steps_per_second": 9.921,
      "step": 6419
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.668513069066991e-06,
      "loss": 5.8524,
      "step": 12838
    },
    {
      "epoch": 2.0,
      "eval_loss": 5.407498359680176,
      "eval_runtime": 64.7053,
      "eval_samples_per_second": 198.407,
      "eval_steps_per_second": 9.922,
      "step": 12838
    }
  ],
  "max_steps": 32095,
  "num_train_epochs": 5,
  "total_flos": 1.689540838293504e+16,
  "trial_name": null,
  "trial_params": null
}