llama.cpp
ggml-cpu: add mxfp4 VSX intrinsics for Power9+ (ppc64le) hardware – 3-4x performance boost
#15385
Merged

Loading