Training Compute-Optimal Large Language Models by from Hacker News on 2022-04-02 22:08 (#5XSK6) Comments