llama.cpp

server: Add "tokens per second" information in the backend

#10548

Merged