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

fault-tolerant: fix resetting "current" progress at the end of successful epoch #8837

Closed
wants to merge 1 commit into from

Conversation

awaelchli
Copy link
Contributor

What does this PR do?

Fixes #8835

Does your PR introduce any breaking changes? If yes, please list them.

Before submitting

  • Was this discussed/approved via a GitHub issue? (not for typos and docs)
  • Did you read the contributor guideline, Pull Request section?
  • Did you make sure your PR does only one thing, instead of bundling different changes together?
  • Did you make sure to update the documentation with your changes? (if necessary)
  • Did you write any new necessary tests? (not for typos and docs)
  • Did you verify new and existing tests pass locally with your changes?
  • Did you list all the breaking changes introduced by this pull request?
  • Did you update the CHANGELOG? (not for typos, docs, test updates, or internal minor changes/refactorings)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:

  • Is this pull request ready for review? (if not, please submit in draft mode)
  • Check that all items from Before submitting are resolved
  • Make sure the title is self-explanatory and the description concisely explains the PR
  • Add labels and milestones (and optionally projects) to the PR so it can be classified

Did you have fun?

Make sure you had fun coding 🙃

Comment on lines +232 to +234
self.batch_progress.current.reset()
self.scheduler_progress.current.reset()
self.batch_loop.optim_progress.reset_on_epoch()
Copy link
Contributor Author

Choose a reason for hiding this comment

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

same as in reset() above line 90-93

Copy link
Contributor

Choose a reason for hiding this comment

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

Probably worth extracting into a protected method

@awaelchli awaelchli added the bug Something isn't working label Aug 10, 2021
@codecov
Copy link

codecov bot commented Aug 10, 2021

Codecov Report

Merging #8837 (f69dbc4) into master (3096ab8) will decrease coverage by 45%.
The diff coverage is 0%.

@@           Coverage Diff            @@
##           master   #8837     +/-   ##
========================================
- Coverage      89%     43%    -45%     
========================================
  Files         169     169             
  Lines       14072   14072             
========================================
- Hits        12466    6117   -6349     
- Misses       1606    7955   +6349     

@@ -229,6 +229,10 @@ def on_run_end(self) -> List[List[STEP_OUTPUT]]:

self.update_lr_schedulers("epoch", update_plateau_schedulers=True)

self.batch_progress.current.reset()
self.scheduler_progress.current.reset()
self.batch_loop.optim_progress.reset_on_epoch()
Copy link
Contributor

Choose a reason for hiding this comment

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

Is it possible to move batch_loop.optim_progress reset to its own loop ?

Copy link
Contributor

Choose a reason for hiding this comment

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

Notice that it's calling reset_on_epoch, not reset. This loop is the epoch loop

Copy link
Contributor

@carmocca carmocca left a comment

Choose a reason for hiding this comment

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

Does the bug happen because restarting=False and the current values correspond to those of the finished epoch? So we never roll over to the next one?

@awaelchli awaelchli closed this Aug 19, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

Successfully merging this pull request may close these issues.

resume_from_checkpoint broken when fault-tolerant feature enabled
3 participants