exllamav2. A fast inference library for running LLMs locally on modern consumer-class GPUs

github.com/19h/exllamav2
by @19h

Problem

Audience

Keep exploring

Updates

No founder updates yet. Check back soon.