tanhuajie2001

/

Reason-RFT-Zero-Visual-Counting-Qwen2-VL-7B

like 0

This PR improves the model card for the Reason-RFT model by:

• Adding pipeline_tag: image-text-to-text for better discoverability on the Hub, as it is a Vision-Language Model for visual reasoning.
• Adding library_name: transformers, as evidenced by the model's architecture (Qwen2VLForConditionalGeneration) and processor/tokenizer classes (Qwen2VLProcessor, Qwen2Tokenizer) found in config.json and tokenizer_config.json. This will enable the automated "How to use with Transformers" widget.
• Updating the main title from "Reason-RFT CoT Dataset" to "Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning" to accurately reflect that this is a model card.
• Adding a direct link to the Hugging Face paper page: Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models.
• Synchronizing the "RoboBrain" mention to reflect "RoboBrain 2.0" with its updated link, matching the project's GitHub README.
• Updating the "News" section to include the most recent announcements from the project's GitHub README.
• Changing the "Model List" section title to "Model Zoo" for consistency with the GitHub README.
• Updating the "Citation" section to include all relevant BibTeX entries from the project's GitHub README.

Please review and merge this PR.