Improve model card: Add detailed framework and results sections

#2
by nielsr HF Staff - opened

This PR enhances the model card by integrating more comprehensive details directly from the project's GitHub repository.

Key additions include:

  • A detailed "Overview" section explaining the problem, the UFO framework solution, and its impact.
  • A dedicated "UFO Framework Details" section outlining the problem formulation, how Unary Feedback as Observation (UFO) works, the training approach with PPO, and the reward design strategies.
  • A "Key Results" section presenting the multi-turn reasoning performance, effectiveness of unary feedback, and the impact of reward design, complete with supporting figures from the paper's project page.
  • An "Acknowledgements" section to properly credit contributing teams.

These additions significantly enrich the model card, providing a more thorough understanding of the model's capabilities, training methodology, and empirical performance.

LichengLiu03 changed pull request status to merged

Sign up or log in to comment