jusjinuk

/

Llama-2-70b-hf-3bit-GuidedQuant-QTIP

like 0

This PR improves the model card for the Llama-2-70b-hf-3bit-GuidedQuant-QTIP model by adding key metadata and enhancing its content:

• Adds pipeline_tag: text-generation for better discoverability and to enable the text generation inference widget.
• Adds library_name: transformers to correctly identify the library used for loading and interacting with the model, ensuring better integration with the Hugging Face ecosystem.
• Corrects the license from llama2 to mit, aligning with the explicit license stated in the original GuidedQuant GitHub repository for the project's artifacts.
• Updates the paper link to the official Hugging Face Papers page: GuidedQuant: Large Language Model Quantization via Exploiting End Loss Guidance.
• Adds a direct link to the project page: https://jusjinuk.me/blog/guidedquant/.
• Includes a clear Python usage example for quick inference, derived from the project's quick start guide.