sunblaze-ucb

/

Qwen2.5-1.5B-GRPO-MATH-1EPOCH

like 0

Follow

sunblaze-ucb 24

This PR significantly enhances the model card for Qwen2.5-1.5B-GRPO-MATH-1EPOCH by integrating key information from its foundational paper and GitHub repository.

The updates include:

• Adding the library_name: transformers metadata tag, which enables the "how to use" widget on the model page, making it easier for users to get started.
• Providing a direct link to the official GitHub repository, granting easy access to the project's code and training scripts.
• Expanding the model description with the paper's abstract and an overview of the "Reinforcement Learning from Internal Feedback (RLIF)" framework, including illustrative figures.
• Incorporating benchmark results to showcase the model's performance and generalization capabilities.
• Adding a practical Python code snippet for sample usage, guiding users on how to run inference with the model.

These improvements will make the model more discoverable, understandable, and user-friendly on the Hugging Face Hub.