How to optimize LLMs for accuracy (via shane) — discussion

#ai