trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 4.2080000000000004e-05,
      "loss": 2.8303,
      "step": 100
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.408e-05,
      "loss": 2.6258,
      "step": 200
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.6079999999999998e-05,
      "loss": 2.5268,
      "step": 300
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.808e-05,
      "loss": 2.4828,
      "step": 400
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.008e-05,
      "loss": 2.4437,
      "step": 500
    }
  ],
  "max_steps": 625,
  "num_train_epochs": 1,
  "total_flos": 4.3203277357056e+16,
  "trial_name": null,
  "trial_params": null
}