llama.cpp

537cdd3a - add server bench script docker