sunblaze-ucb

/

Qwen2.5-3B-GRPO-MATH-1EPOCH

like 0

Follow

sunblaze-ucb 38

This PR enhances the model card for Qwen2.5-3B-GRPO-MATH-1EPOCH by:

• Adding library_name: transformers to the metadata, enabling the "Use in Transformers" button and improving integration.
• Including specific tags like mathematical-reasoning, code-generation, reinforcement-learning, and reasoning for better discoverability and categorization.
• Updating the model description to provide comprehensive context from the paper's abstract about the "Intuitor" project and "Reinforcement Learning from Internal Feedback (RLIF)".
• Adding a direct link to the official GitHub repository for easy access to the code.
• Providing a practical Python code snippet for sample usage with the transformers library, demonstrating how to load and perform text generation, particularly for mathematical reasoning.
• Explicitly linking to the paper: Learning to Reason without External Rewards.

These updates will make the model more accessible, informative, and aligned with Hugging Face's model card best practices.