PaDT-MLLM

/

PaDT_REC_7B

like 0

This PR enhances the model card for Patch-as-Decodable Token (PaDT) by:

1. Adding library_name: transformers, which enables direct integration and automated code snippets via the Hugging Face transformers library, as evidenced by the sample usage.
2. Refining the pipeline_tag into a list pipeline_tags including both image-text-to-text and image-segmentation. This accurately reflects the model's capabilities to generate both textual and diverse visual outputs, including segmentation masks, as described in the paper abstract and showcase examples.
3. Updating the introductory sentence for the sample usage section to match the GitHub README for better context.

These changes will improve the model's discoverability on the Hub and provide clearer guidance for users.