llama.cpp

Add quantization recipes from custom recipe files

#21070

Open