trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.3133208255159474,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38,
      "learning_rate": 9.628378378378378e-07,
      "loss": 4.3599,
      "step": 1000
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.253003003003003e-07,
      "loss": 2.1723,
      "step": 2000
    },
    {
      "epoch": 1.13,
      "learning_rate": 8.877627627627627e-07,
      "loss": 0.7581,
      "step": 3000
    }
  ],
  "max_steps": 26650,
  "num_train_epochs": 10,
  "total_flos": 5.029954774155264e+17,
  "trial_name": null,
  "trial_params": null
}