llama.cpp

update convert-ggml-to-pth.py

#686

Merged