Logics-MLLM

/

Logics-Parsing

like 47

Follow

Logics-MLLM 26

This PR adds a comprehensive model card for the Logics-Parsing model. It includes:

• Linking the model to its paper: Logics-Parsing: End-to-End LVLM for Document Parsing.
• Adding apache-2.0 as the license (already present).
• Setting library_name: transformers to enable the automated "how to use" widget, justified by the model's architecture files (Qwen2_5_VLForConditionalGeneration, Qwen2_5_VLProcessor, Qwen2Tokenizer) and acknowledgments to Qwen2.5-VL.
• Specifying pipeline_tag: image-text-to-text for better discoverability, as it's a Vision-Language Model for document parsing.
• Adding relevant tags: document-parsing, vlm, qwen.
• Including links to the GitHub repository and ModelScope demo.
• Incorporating an introduction, key features, benchmark results, and a shell command snippet for quick inference, all sourced directly from the GitHub README.
• Correctly formatted the citation block for the paper.

Please review and merge if everything looks good.