Fsfairx Gemma2 RM V0.1

🚀 Vanilla BT-based Reward Model with Gemma-2-9B

This project presents a Vanilla BT based Reward model built upon Gemma-2-9B. The recipes are sourced from the RLHF Workflow, offering a reliable approach for reward modeling.

🚀 Quick Start

The model has been evaluated on the reward-bench, and here are the results:

📚 Documentation

For more information about the RLHF Workflow, please refer to the following paper:

@misc{dong2024rlhf,
      title={RLHF Workflow: From Reward Modeling to Online RLHF}, 
      author={Hanze Dong and Wei Xiong and Bo Pang and Haoxiang Wang and Han Zhao and Yingbo Zhou and Nan Jiang and Doyen Sahoo and Caiming Xiong and Tong Zhang},
      year={2024},
      eprint={2405.07863},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}