llama.cpp
CUDA: fix replacment of bad archs in CMake
#18457
Merged

Loading