Hacker News new | ask | show | jobs
by sgt101 350 days ago
LLMs are deterministic.

If you run an LLM with optimization turned on on a NVIDIA GPU then you can get non-deterministic results.

But, this is a choice.