3 ms·

I mean they already supported the project quite a bit. @ngxson and maybe others? from Huggingface are big contributors to llama.cpp.