{
    "reward_model": {
        "model_path": "parth-ptl-97/reinforcement-learning-human-feedback"
    },
  "model_type":"gpt2"
}