UserJoseph

/

DisTime-1B

like 0

This PR significantly enhances the model card by:

• Adding the pipeline_tag: video-text-to-text, allowing the model to be discovered under relevant filters on the Hub.
• Specifying library_name: transformers, enabling the "How to use" widget for easier inference.
• Adding relevant tags such as multimodal, video-understanding, temporal-localization, and qwen for improved discoverability and context.
• Linking directly to the Hugging Face paper page: DisTime: Distribution-based Time Representation for Video Large Language Models.
• Providing a link to the official GitHub repository for code and further details.
• Including the full abstract and a clear transformers-based usage example for quick understanding and implementation.
• Adding the citation information and acknowledgements.
• Removing the unnecessary "File information" section.