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

Include timestamp and span-compatible attributes in done files #23

Merged
merged 3 commits into from
Dec 3, 2024

Conversation

meatballhat
Copy link
Contributor

@meatballhat meatballhat commented Dec 3, 2024

so that we can pass along more context through director and generally depend on more than an empty file.

Please note: this also includes a behavior change to is_done that no longer deletes the directory tree when .done is not present. I couldn't come up with a reason for why we would want this outside of testing. I'm happy to learn that it's important and we should keep it that way 😇 (although if it needs to be kept that way then I'd like it to have a different name).

Update: What was previously monobase.util:is_done is now monobase.util:require_done_or_rm

Connected to PLAT-555

so that we can pass along more context through `director` and generally
depend on more than an empty file.

Connected to PLAT-555
@meatballhat meatballhat requested a review from a team as a code owner December 3, 2024 04:05
Copy link
Contributor

@nevillelyh nevillelyh left a comment

Choose a reason for hiding this comment

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

We delete a directory when .done is missing so that a step can be retried if it failed previously, e.g. OOM when running CUDA runfile, network issues, yanked torch nightly etc., all of which has happened in prod. Otherwise the directory might be in a corrupted or incomplete state.

@meatballhat
Copy link
Contributor Author

@nevillelyh Thank you. I'll propose a rework ⌨️

@nevillelyh
Copy link
Contributor

FWIW everything else looks good, should be good if we just reinstate the rmtree.

@meatballhat meatballhat merged commit f205e31 into main Dec 3, 2024
3 checks passed
@meatballhat meatballhat deleted the done-contains-stuff branch December 3, 2024 16:34
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