I stress-tested LLM quantization by deliberately breaking models

(github.com) by GraceJackson07 | Sep 23, 2026 | 0 comments on HN