llama.cpp

ggml WebGPU: add support for quantization types

#15440

Merged