Smp grad accum #10488
Merged
Smp grad accum #10488
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
commented
Mar 2, 2021
| @@ -108,7 +107,7 @@ def _wrap_model(self, model, training=True): | |||
| # Wrapping the base model twice in a DistributedModel will raise an error. | |||
| if isinstance(self.model_wrapped, smp.model.DistributedModel): | |||
| return self.model_wrapped | |||
| return smp.DistributedModel(model) | |||
| return smp.DistributedModel(model, backward_passes_per_step=self.args.gradient_accumulation_steps) | |||
There was a problem hiding this comment.
This does the equivalent of no_sync in regular DDP.
Comment on lines
+741
to
+745
| def _no_sync_in_gradient_accumulation(self): | ||
| """ | ||
| Whether or not to use no_sync for the gradients when doing gradient accumulation. | ||
| """ | ||
| return not self.deepspeed |
There was a problem hiding this comment.
This is introduced to make it easy to skip the no_sync part in subclasses.
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Outdated suggestions cannot be applied.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 adds support for gradient accumulation in
SageMakerTrainer. It has been tested on the glue script with success (with and without gradient accumulation passed along).