InternRobotics

/

InternVLA-M1-LIBERO-Spatial

like 11

Follow

Intern Robotics 221

This PR significantly enhances the model card for InternVLA-M1_spatial by:

• Adding pipeline_tag: robotics to improve discoverability on the Hugging Face Hub.
• Specifying library_name: transformers, as the model's codebase is built upon both Transformers and Diffusers, with its VLM component (Qwen2.5-VL) being a Transformers model.
• Integrating the official Hugging Face paper link: InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy.
• Including the paper's abstract for a quick overview of the model's capabilities and methodology.
• Adding a comprehensive "Quick Interactive M1 Demo" section with Python code snippets for both chat/spatial grounding and action prediction, directly from the GitHub repository, enabling users to quickly get started.
• Incorporating additional detailed sections from the GitHub README such as Key Features, Target Audience, Experimental Results, Model Zoo, Roadmap, Contributing, Contact, and Acknowledgements, providing a more complete resource.
• Updating the citation with a more complete BibTeX entry.

These updates aim to make the model card more informative, user-friendly, and discoverable.

Please review and merge this PR.