Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix logging verbosity in HF model download script and repair symlinks #727

Merged
merged 5 commits into from
Nov 10, 2023

Conversation

jerrychen109
Copy link
Contributor

@jerrychen109 jerrychen109 commented Nov 9, 2023

This PR temporarily suppresses InvalidCertWarning when downloading a HF model from a cache server with ignore_cert=True, as the warnings are quite noisy. In addition, it also sets the default log level to INFO in the download script so the logging is actually visible.

Also modifies the script to call download_from_hf_hub even after a successful cache download for the purpose of repairing the Hugging Face symlinks to address a non-deterministic bug where the symlinks aren't properly repaired by the time composer loads the model.

Tested a run downloading MPT-7B from a cache server using the new commit 94eff5bae6210865e0a7ee60fe2ec7eba4750522. Note the no-op for Fetching 24 files resulting from the download from HF after:
image

And successful model initialization:
image

@jerrychen109 jerrychen109 changed the title Fix logging verbosity in HF model download script Fix logging verbosity in HF model download script and repair symlinks Nov 9, 2023
@jerrychen109 jerrychen109 marked this pull request as draft November 9, 2023 22:20
@jerrychen109 jerrychen109 marked this pull request as ready for review November 9, 2023 22:59
Copy link
Collaborator

@dakinggg dakinggg left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, thanks!

@jerrychen109 jerrychen109 enabled auto-merge (squash) November 10, 2023 00:59
@jerrychen109 jerrychen109 merged commit 7c4d24a into main Nov 10, 2023
12 checks passed
@dakinggg dakinggg deleted the jerry/model-download-util-logs branch November 17, 2023 06:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants