llama.cpp
nix: ci: dont test cuda and rocm (for now)
#6349
Merged

Loading