llama.cpp

server bench: fix bench not waiting for model load

#7284

Merged