Improve model card for Lego-Edit: Add metadata, links, abstract, and usage guidance
#1
by
nielsr
HF Staff
- opened
This PR significantly enhances the model card for Lego-Edit by:
- Updating the license: Changed
licensefromapache-2.0tocc-by-nc-4.0to accurately reflect the model's licensing as stated in the GitHub repository's disclaimer. - Adding metadata: Included
library_name: transformersto enable automated code snippets, as evidenced by the model's architecture (Qwen2_5_VLForConditionalGeneration) and setup instructions in the GitHub README. Also addedpipeline_tag: image-to-imageto improve discoverability for relevant tasks. - Providing comprehensive links: Added direct links to the paper (Hugging Face Papers and arXiv), the project page (https://xiaomi-research.github.io/lego-edit/), the GitHub repository (https://github.com/xiaomi-research/lego-edit), and a live demo (https://editdemo.ai.xiaomi.net/).
- Including abstract and features: Added the paper abstract and the key features of Lego-Edit for a clear overview of the model's capabilities.
- Guiding usage: Directed users to the GitHub repository's "Quick Start" section for detailed environment setup and how to run the Gradio WebUI, adhering to the instruction not to make up code snippets.
- Adding Disclaimer, Citation, and Acknowledgments: Incorporated these important sections from the original GitHub README to provide full context and proper attribution.
These updates aim to provide a more informative and user-friendly experience for anyone interacting with the Lego-Edit model on the Hugging Face Hub.