Qinsi1

/

Vision-Zero-Qwen-2.5-VL-7B-Chart

Like 1

This PR significantly enhances the model card for Vision-Zero-Qwen-2.5-VL-7B-Clevr by adding:

• The pipeline_tag: image-text-to-text to improve discoverability and categorize the model's functionality.
• The library_name: transformers metadata, as evidenced by config.json, enabling the automated "Use in Transformers" widget.
• The license: cc-by-nc-4.0 as a common research artifact license.
• The base_model: Qwen/Qwen2.5-VL-7B for clarity on its foundation.
• The comprehensive abstract from the paper.
• A direct link to the paper: Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play.
• A link to the official GitHub repository: https://github.com/wangqinsi1/Vision-Zero.
• An overview image sourced from the GitHub repository.
• A table listing Vision-Zero models and datasets, as provided in the GitHub README.
• A detailed "Sample Usage" section, including a Python code snippet for inference, directly copied from the GitHub README.
• The BibTeX citation information.

These updates aim to provide clearer and more accessible information for users and researchers.