llama.cpp

ggml: aarch64: implement SVE kernels for q4_K_q8_K vector dot

#11227

Merged