xiaomi-research

/

lego-edit

like 1

Follow

Xiaomi Research 11

This PR significantly enhances the model card for Lego-Edit, improving its discoverability and providing comprehensive information for users.

Key changes include:

• Updated License: The license metadata has been updated to cc-by-nc-4.0 to accurately reflect the license for the model artifacts, as stated in the GitHub repository's disclaimer.
• Added pipeline_tag: image-to-image has been added to enable better filtering and discovery on the Hugging Face Hub for image editing models.
• Added library_name: transformers has been added due to the model's architecture (Qwen2_5_VL) and processor configurations found in config.json, preprocessor_config.json, and tokenizer_config.json, which are indicative of transformers compatibility.
• Added tags: image-editing, multimodal, and mllm tags have been included to further categorize the model based on its capabilities and architecture.
• Comprehensive Content: The README now includes:
  • Direct links to the paper (Lego-Edit: A General Image Editing Framework with Model-Level Bricks and MLLM Builder), the official project page (https://xiaomi-research.github.io/lego-edit/), and the GitHub repository (https://github.com/xiaomi-research/lego-edit).
  • The abstract of the paper for a quick overview.
  • Detailed sections on model features, quick start instructions (including environment setup and running the Gradio demo), new tool integration, and more usage examples, all directly sourced from the GitHub README.
  • A clear disclaimer regarding licensing, intellectual property, and responsible use.
  • A citation section for proper attribution.
  • Acknowledgments to foundational projects.
• Image Linking: All local image paths have been converted to absolute URLs pointing to the GitHub repository's raw content for correct rendering on the Hub.

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.