Open grencez opened 7 months ago
I'm going to wait for llama.cpp's sampling refactor. It looks like some of the project's CLI is going to be exposed as a library, so the most reliable implementation on our end could just hook into that.
I'm going to wait for llama.cpp's sampling refactor. It looks like some of the project's CLI is going to be exposed as a library, so the most reliable implementation on our end could just hook into that.