llama.cpp
ggml : fix IQ3_S AVX implementation
#5834
Merged

Loading