Comment by satvikpendem

I was looking for this and to even build something like this in order to have a pure Rust stack for my backend, as well as embed local inference on edge like on mobile, but llama.cpp is just too far advanced to catch up, it seems. There are some but I'm not sure how good they are.