Fix lr scheduler not being reset on reruns #24758
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sgugger
approved these changes
Jul 11, 2023
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. |
4 tasks
Lorenzobattistela
pushed a commit
to Lorenzobattistela/transformers
that referenced
this pull request
Jul 13, 2023
* Try this * Solved! * Rm extranious * Rm extranious * self * Args' * Check for if we created the lr scheduler * Move comment * Clean
blbadger
pushed a commit
to blbadger/transformers
that referenced
this pull request
Nov 8, 2023
* Try this * Solved! * Rm extranious * Rm extranious * self * Args' * Check for if we created the lr scheduler * Move comment * Clean
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR ensures that a new learning rate scheduler is created each time we rerun the
inner_training_loop, so that if we have an lr such aslinear, a new LR is generated based on the new batch size and step count. I don't believe a new optimizer is needed here to be recreated, just the scheduler as adjusting the bs and lr shouldn't matter? But if we think it is we can go ahead and add a reset to the optimizer as well.Fixes # (issue)
The true solution to #24521
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@sgugger