DiHe

/

LLR-20B

Like 0

This PR adds a comprehensive model card for the LLR-20B model. It includes:

• Metadata for pipeline_tag (text-generation), library_name (transformers), and license (apache-2.0).
• A description of the Layerwise Learning Rate (LLR) method and the specific checkpoint details.
• Links to the paper and the official GitHub repository for further reference.