llama.cpp
vulkan: work around Intel ANV fp16 bug in MMQ
#18814
Merged

Loading