llama.cpp

ggml-cpu: add 128-bit RVV implementation for Quantization Vector Dot

#20633

Open