Local LLM inference – impressive but too hard to work with by from Hacker News on 2025-04-21 16:42 (#6WRW6) Comments