llama.cpp

2cca6c01 - rpc : add command line option for number of threads for the CPU backend (#13060)