Use model.from_pretrained for DataParallel also #8795
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
When training on multiple GPUs, the code wraps a model with torch.nn.DataParallel. However if the model has custom from_pretrained logic, it does not get applied during load_best_model_at_end. This commit uses the underlying model during load_best_model_at_end, and re-wraps the loaded model with DataParallel. If you choose to reject this change, then could you please move the this logic to a function, e.g. def load_best_model_checkpoint(best_model_checkpoint) or something, so that it can be overridden?
sgugger
reviewed
Nov 30, 2020
src/transformers/trainer.py
Outdated
Comment on lines
817
to
819
| if is_data_parallel: | ||
| # re-wrap with DataParallel | ||
| self.model = torch.nn.DataParallel(self.model) |
There was a problem hiding this comment.
The model attribute of the Learner is never wrapped in DataParallel or the like, it stays a reference to the original model so these lines are not necessary.
With the same reasoning, I think the only change to make is on the line above this comment (L816) and replace
self.model = model.from_pretrained(self.state.best_model_checkpoint)
by
self.model = self.model.from_pretrained(self.state.best_model_checkpoint)
Thanks for the feedback. I made the change that you proposed, but I also think we should update L811 to check if `self.mode` is an instance of `PreTrained`, otherwise we would still not get into that `if` section, right?
|
Oh, looks like there is a last code-style issue to fix. Could you run |
|
I don't have |
|
|
|
Weird indeed. Will merge and fix if the issue persists. |
stas00
pushed a commit
to stas00/transformers
that referenced
this pull request
Dec 5, 2020
* Use model.from_pretrained for DataParallel also When training on multiple GPUs, the code wraps a model with torch.nn.DataParallel. However if the model has custom from_pretrained logic, it does not get applied during load_best_model_at_end. This commit uses the underlying model during load_best_model_at_end, and re-wraps the loaded model with DataParallel. If you choose to reject this change, then could you please move the this logic to a function, e.g. def load_best_model_checkpoint(best_model_checkpoint) or something, so that it can be overridden? * Fix silly bug * Address review comments Thanks for the feedback. I made the change that you proposed, but I also think we should update L811 to check if `self.mode` is an instance of `PreTrained`, otherwise we would still not get into that `if` section, right?
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.
When training on multiple GPUs, the code wraps a model with torch.nn.DataParallel. However if the model has custom from_pretrained logic, it does not get applied during load_best_model_at_end.
This commit uses the underlying model during load_best_model_at_end, and re-wraps the loaded model with DataParallel.
If you choose to reject this change, then could you please move the this logic to a function, e.g. def load_best_model_checkpoint(best_model_checkpoint) or something, so that it can be overridden?
What does this PR do?
Fixes # (issue)
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 which may be interested in your PR.