llama.cpp

gguf-py: byteswapping improvements

#12851

Merged