gguf-py PyPI release question

[PenutChen]

The latest release of gguf-py was on Dec 13, 2023. There are many changes compared to the current source. Most users usually use the source version of gguf-py. Recently, Hugging Face has integrated with the gguf format, but it depends on the PyPI version of gguf-py, which causes some compatibility issues (huggingface/transformers#31725).

I think this makes it hard to decide which version to follow. New gguf models usually come out with the latest gguf-py, but third-party integrations can only use the old package on PyPI. I'm wondering if future versions of gguf-py will be released on PyPI?

[ggerganov]

I tried to run the PyPI workflow, but it no longer works:

https://github.com/ggerganov/llama.cpp/actions/runs/9820095645/job/27114309369#step:7:30

Uploading gguf-0.9.0-py3-none-any.whl
WARNING  Error during upload. Retry with the --verbose option for more details. 
ERROR    HTTPError: 400 Bad Request from https://upload.pypi.org/legacy/        
         User 'monatis' does not have a verified primary email address. Please  
         add a verified primary email before attempting to upload to PyPI. See  
         https://pypi.org/help/#verified-email for more information.      

Pinging @monatis

[monatis]

Thanks @ggerganov will have a look at this in the morning.

[mgoin]

Hey @monatis thanks for looking into this, any update to share?

[monatis]

Verified email and enabled TFA. Will test the pipeline now.

[monatis]

Seems like my permissions purged --cannot make a push.

remote: Permission to ggerganov/llama.cpp.git denied to monatis.                                                        
fatal: unable to access 'https://github.com/ggerganov/llama.cpp.git/': The requested URL returned error: 403            

[ggerganov]

You are still a collaborator to the project - should be able to push

[monatis]

Ok needed to update credentials on this device. PR #8410