llama.cpp

server: implement /models?reload=1

#21848

Merged