llama.cpp

drop quantize.py (now that models are using a single file)

#640

Merged