logoalt Hacker News

FairyFuse: Multiplication-Free LLM Inference on CPUs via Fused Ternary Kernels

5 pointsby PaulHouletoday at 5:35 PM0 commentsview on HN

Comments