OpenMMReasoner

/

OpenMMReasoner-RL

like 5

Follow

OpenMMReasoner 4

This PR significantly enhances the model card for OpenMMReasoner by adding essential metadata and incorporating detailed content from the project's GitHub README.

Metadata Additions:

• license: apache-2.0: Based on common open-source practices for similar projects and the majority consensus from colleagues.
• pipeline_tag: image-text-to-text: Accurately categorizes the model as a multimodal reasoning model that processes images and text inputs to generate text. This improves discoverability on the Hub.
• library_name: transformers: Confirmed by the transformers imports in the "Basic Usage" code snippet, enabling the automated "How to use" widget on the model page.

Content Additions:

• Evaluation Results: The comprehensive evaluation table and accompanying text, detailing the model's performance across various benchmarks, have been integrated.
• Citation: The BibTeX entry for the paper has been added, allowing researchers to easily cite the work.
• Acknowledgements: The section acknowledging the underlying open-source projects has been included.

These changes provide a more complete, discoverable, and user-friendly model card for the OpenMMReasoner model.