LCO-Embedding

/

LCO-Embedding-Omni-7B

like 3

Follow

LCO-Embedding 7

This PR significantly enhances the model card for the LCO-Embedding model by adding essential metadata and a more detailed description of its capabilities.

Specifically:

• Adds pipeline_tag: feature-extraction to improve discoverability on the Hugging Face Hub, as this model is designed for omnimodal representation learning.
• Adds library_name: transformers, indicating compatibility with the Hugging Face Transformers library, as evidenced by the config.json and tokenizer_config.json files.
• Updates the content to provide a comprehensive overview of the model, including its key contributions (LCO-Embedding framework, Generation-Representation Scaling Law, SeaDoc task), and incorporates evaluation results with illustrative figures directly from the GitHub repository.
• Explicitly links to the paper and the GitHub repository for easy access to more information and code.

Please review and merge this PR if everything looks good.