[Trainer] Add nan/inf logging filter #13619
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
Sep 17, 2021
There was a problem hiding this comment.
Thanks for adding this! I wonder if the default of the new feature shouldn't be True? It's a little bi breaking, but only for the logging, not the actual training, and it makes more sense to ignore thoses NaN/infs (which are ignored by default by the optimizer since there is some gradient clipping).
src/transformers/training_args.py
Outdated
Comment on lines
475
to
476
| "Filter nan and inf losses for logging. " | ||
| "Note this flag only affects the logging output and not the optimization step." |
There was a problem hiding this comment.
No need to be this long here, but there should be a proper documentation above that can be longer!
|
Happy to let it default to True! |
| @@ -468,6 +478,7 @@ class TrainingArguments: | |||
| ) | |||
| logging_first_step: bool = field(default=False, metadata={"help": "Log the first global_step"}) | |||
| logging_steps: int = field(default=500, metadata={"help": "Log every X updates steps."}) | |||
| logging_nan_inf_filter: str = field(default=True, metadata={"help": "Filter nan and inf losses for logging."}) | |||
There was a problem hiding this comment.
The default is now True
Albertobegue
pushed a commit
to Albertobegue/transformers
that referenced
this pull request
Jan 13, 2022
* finish * add test * push * remove unnecessary code * up * correct test * Update src/transformers/training_args.py
Albertobegue
pushed a commit
to Albertobegue/transformers
that referenced
this pull request
Jan 27, 2022
* finish * add test * push * remove unnecessary code * up * correct test * Update src/transformers/training_args.py
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?
Some losses sometimes produce
inflosses which however doesn't necessarily mean that the training went bad. When using the CTC loss for speech recognition - see: #13620 this is often the case. The problem is that as soon as one single loss step isinfornanthe rest of the training logs will displayinfornan. In this PR a flag is added that allows the user to filter outnanandinfvalues for training. It defaults toFalseand will be set toTruein all CTC training scripts.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.