Comment by tempest_

I use CC, and I understand what caching means.

I have no idea how that works with a LLM implementation nor do I actually know what they are caching in this context.