craigwu

/

visual_jigsaw_video_7B

like 1

This PR adds a comprehensive model card for the Visual Jigsaw Video 7B model, based on the paper Visual Jigsaw Post-Training Improves MLLMs.

The updates include:

• Relevant metadata: license (apache-2.0), pipeline_tag (video-text-to-text), library_name (transformers), base_model (Qwen2.5-VL-7B-Instruct), and additional tags.
• A link to the paper on Hugging Face Papers.
• Links to the project page and the GitHub repository.
• A concise overview of the model and its capabilities, including the overview image.
• A sample usage section demonstrating how to use the model with the transformers library, building upon the original repository's instructions regarding its Qwen2.5-VL-7B-Instruct base.
• The paper's BibTeX citation.

This should significantly improve the discoverability and usability of the model on the Hub.