Replies: 1 comment 1 reply
-
You would use it like this: save_dict = {"test": mx.ones((4, 4), dtype=mx.int32)}
mx.save_gguf("array.gguf", save_dict) |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi all! Unfortunately I'm struggling again a bit with the docs; I can't find a detailed enough explanation for how to use the mlx.core.save_gguf command so I'm confused as to how to convert to a .gguf. I did try GPT-4 but alas, it was of no help. Would really love to be able to convert to a 4 or even 8 bit .gguf file so I hope this is possible :) - if it's with a different command please let me know!
Thanks everyone 🙌
Beta Was this translation helpful? Give feedback.
All reactions