husj576

/

GRIFFIN-llama3-instruct-70B

like 0

This PR significantly enhances the model card for husj576/GRIFFIN-llama3-instruct-8B by:

• Adding pipeline_tag: text-generation for better discoverability on the Hugging Face Hub.
• Adding library_name: transformers as the model's architecture is compatible with the Transformers library.
• Linking directly to the associated paper: GRIFFIN: Effective Token Alignment for Faster Speculative Decoding.
• Providing a direct link to the GitHub repository: https://github.com/hsj576/GRIFFIN.
• Populating the content section with the paper's abstract, key features from the GitHub README, performance benchmarks (images), and a detailed Python code snippet for inference, adapted from the original repository for clarity with Hugging Face model IDs.

This ensures that users have comprehensive information and a clear path to utilize the model effectively.