llama.cpp
vulkan: initial support for IQ4_XS quantization
#11501
Merged

Loading