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

github.com/oobabooga/exllamav2

Problem

Audience

Updates

No founder updates yet. Check back soon.