ref:
ggerganov/ggml#302
#1991
This PR paves the way for integrating more models into llama.cpp. It changes the file format in which we convert the models by extending it with key-value pairs meta...
They still consider it a beta but there we go! It’s happening :D
AFAIK supposedly GGUF is a more extensible format that contains (or can contain) more metadata types that make it usable for different model architectures. The main advantage is that this should be the last breaking format change, as future changes can be added in a more modular way.
AFAIK supposedly GGUF is a more extensible format that contains (or can contain) more metadata types that make it usable for different model architectures. The main advantage is that this should be the last breaking format change, as future changes can be added in a more modular way.