Fix no-grad grad-fn lookup in ZeRO hook counting on PyTorch 2.3 (#7830)#7841
Open
tohtana wants to merge 2 commits intodeepspeedai:masterfrom
Open
Fix no-grad grad-fn lookup in ZeRO hook counting on PyTorch 2.3 (#7830)#7841tohtana wants to merge 2 commits intodeepspeedai:masterfrom
tohtana wants to merge 2 commits intodeepspeedai:masterfrom
Conversation
Signed-off-by: Masahiro Tanaka <mtanaka@anyscale.com>
PyTorch 2.3's _get_grad_fn_or_grad_acc can fail in backward hooks because it calls view_as() without torch.enable_grad(). DeepSpeed invokes this from count_used_parameters_in_backward(), so ZeRO stage 1/2/3 can hit AttributeError: 'NoneType' object has no attribute 'next_functions'. Wrap the lookup in torch.enable_grad() to match newer PyTorch behavior and keep hook counting semantics intact. Update the unit test to validate grad mode is enabled for this lookup path instead of swallowing AttributeError. Signed-off-by: Masahiro Tanaka <mtanaka@anyscale.com>
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #7830
In torch==2.3,
_get_grad_fn_or_grad_acccan fail in backward-hook no-grad context (NoneType.next_functions), which breakscount_used_parameters_in_backward.This PR wraps that lookup with
torch.enable_grad()(matching newer torch behavior) and updates the unit test accordingly.