trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0032038446135362435,
  "eval_steps": 500,
  "global_step": 6,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.00018660254037844388,
      "loss": 2.8052,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.1675,
      "step": 2
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0001,
      "loss": 2.3986,
      "step": 3
    },
    {
      "epoch": 0.0,
      "learning_rate": 5.000000000000002e-05,
      "loss": 2.0054,
      "step": 4
    },
    {
      "epoch": 0.0,
      "learning_rate": 1.339745962155613e-05,
      "loss": 2.2608,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0,
      "loss": 2.1385,
      "step": 6
    }
  ],
  "logging_steps": 1,
  "max_steps": 6,
  "num_train_epochs": 1,
  "save_steps": 10,
  "total_flos": 1000182195486720.0,
  "trial_name": null,
  "trial_params": null
}