tomg-group-umd

/

LoRI-S_nlu_llama3_rank_32

like 0

Follow

Tom Goldstein's Lab at University of Maryland, College Park 181

This PR significantly enhances the model card for LoRI-S_nlu_llama3_rank_32 by:

• Adding the license: apache-2.0 to the metadata.
• Including relevant tags such as lora, peft, nlu, code-generation, mathematical-reasoning, safety-alignment, and multi-task-learning to improve discoverability.
• Updating the main paper link to the Hugging Face Papers page (https://huggingface.co/papers/2504.07448).
• Providing a comprehensive description of the LoRI model, its features, and how it addresses cross-task interference.
• Adding direct links to the official GitHub repository (https://github.com/juzhengz/LoRI/) and the project page (https://juzhengz.github.io/).
• Including a clear Python code snippet for quick inference, making it easier for users to get started.
• Populating various sections such as Model Details, Uses, Bias, Risks, and Limitations, Training Details, and Evaluation with information extracted from the paper abstract and the GitHub repository's README.
• Adding the academic citation for the paper.
• Embedding an illustrative image from the GitHub repository.

These improvements make the model card much more informative, discoverable, and user-friendly for researchers and practitioners.