llama.cpp

gguf-py: allow converting multi-tensor models from read-only locations

#18100

Merged