trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9761904761904763,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6,
      "learning_rate": 3.5594713656387666e-05,
      "loss": 0.9722,
      "step": 200
    },
    {
      "epoch": 1.19,
      "learning_rate": 2.6784140969163e-05,
      "loss": 0.7369,
      "step": 400
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.7973568281938328e-05,
      "loss": 0.6069,
      "step": 600
    },
    {
      "epoch": 2.38,
      "learning_rate": 9.162995594713657e-06,
      "loss": 0.4067,
      "step": 800
    },
    {
      "epoch": 2.98,
      "learning_rate": 3.5242290748898686e-07,
      "loss": 0.3301,
      "step": 1000
    }
  ],
  "max_steps": 1008,
  "num_train_epochs": 3,
  "total_flos": 394670126592000.0,
  "trial_name": null,
  "trial_params": null
}