llama.cpp

docker : fix missing binaries in full-cuda image

#9278

Merged