Skip to content

Implement KV-cache #7

@vdyma

Description

@vdyma

Implement KV-caching variant such as Multi-Query Attention or Grouped-Query Attention. According to LLaMA-2, GQA performs slightly better than MQA.

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions