llama.cpp

ggml: aarch64: implement SVE kernels for q8_0_q8_0, q4_0_q8_0 vector dot

#7433

Merged