Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
Make it possible to skip large files when doing push_to_hub
See original GitHub issueIs your feature request related to a problem? Please describe.
Suppose you have a repo on the hub with some large model files (>1 GB). Suppose that you want to upload some very small tokenizer files to that repo (this was my use case).
Currently when doing
tokenizer.push_to_hub(repo_url="...")
, it will first clone the entire repository (this takes a while!), downloading the big model.pt file, before I’m able to upload my tiny files to the repo.
Describe the solution you’d like
We recently added a kwarg to a Repository called skip_lfs_files, to skip LFS files when cloning a repo (see #472). It would be great to pass this argument when pushing to the hub, something like:
tokenizer.push_to_hub(repo_url="...", skip_lfs_files=True)
Or skip those files by default?
Issue Analytics
- State:
- Created 2 years ago
- Comments:6 (6 by maintainers)
Top Results From Across the Web
Top Related Medium Post
Top Related StackOverflow Question
Troubleshoot Live Code
Top Related Reddit Thread
Top Related Hackernoon Post
Top Related Tweet
Top Related Dev.to Post
Top Related Hashnode Post
As discussed offline this is very linked to #321, which I believe Zach wants to work on next week.
WDYT @muellerzr?
cc @LysandreJik