trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 16.666666666666668,
  "eval_steps": 500,
  "global_step": 250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.33,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.6829,
      "step": 50
    },
    {
      "epoch": 6.67,
      "learning_rate": 0.00013090169943749476,
      "loss": 0.0928,
      "step": 100
    },
    {
      "epoch": 10.0,
      "learning_rate": 6.909830056250527e-05,
      "loss": 0.0746,
      "step": 150
    },
    {
      "epoch": 13.33,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 0.0672,
      "step": 200
    },
    {
      "epoch": 16.67,
      "learning_rate": 0.0,
      "loss": 0.0638,
      "step": 250
    }
  ],
  "logging_steps": 50,
  "max_steps": 250,
  "num_train_epochs": 17,
  "save_steps": 500,
  "total_flos": 153430498934784.0,
  "trial_name": null,
  "trial_params": null
}