checkpoint-1000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3022974607013301,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15,
      "learning_rate": 0.0009848851269649336,
      "loss": 0.3001,
      "step": 500
    },
    {
      "epoch": 0.15,
      "eval_loss": 0.0005047954618930817,
      "eval_runtime": 82.2643,
      "eval_samples_per_second": 50.933,
      "eval_steps_per_second": 3.185,
      "step": 500
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.000969770253929867,
      "loss": 0.1201,
      "step": 1000
    },
    {
      "epoch": 0.3,
      "eval_loss": 0.0015803328715264797,
      "eval_runtime": 82.0816,
      "eval_samples_per_second": 51.047,
      "eval_steps_per_second": 3.192,
      "step": 1000
    }
  ],
  "max_steps": 33080,
  "num_train_epochs": 10,
  "total_flos": 4892330557440000.0,
  "trial_name": null,
  "trial_params": null
}