llama.cpp

94c5652f - quantize: make output filename optional, default to ggml-model-<ftype>.bin (#1301)