ashun989

/

GlimpsePrune_LLaVA-1.5-7B

like 0

This PR significantly enhances the model card for GlimpsePrune by:

• Adding pipeline_tag: image-text-to-text: This improves discoverability on the Hugging Face Hub, allowing users to find the model under relevant Vision-Language tasks (e.g., at https://huggingface.co/models?pipeline_tag=image-text-to-text).
• Specifying library_name: transformers: This indicates compatibility with the Hugging Face Transformers library, enabling the "Use in Transformers" snippet button and facilitating easy integration for users.
• Linking to the paper: The model card now includes a direct link to the paper "A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models" on Hugging Face Papers.
• Including the GitHub repository link: A link to the official code repository https://github.com/HVision-NKU/GlimpsePrune is now prominent.
• Adding detailed model description: Key information from the paper abstract and the project's GitHub README, including a general overview, key features, framework overview, and performance results, has been integrated. This provides a clear understanding of the model's purpose, capabilities, and technical details.
• Providing a clear usage example: A Python code snippet demonstrating how to use the model with the Hugging Face transformers library for inference is included, making it easier for users to get started.
• Embedding illustrative images: Key images from the project's GitHub repository are embedded to visually explain the framework and showcase results.
• Including citation information: The BibTeX entry for the paper is added for proper attribution.

This PR aims to make the GlimpsePrune model more accessible and informative for the community.