Fix FSDP gradient calculation with orig params #9335
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
|
Nice, can you fix the DCO? |
akoumpa
approved these changes
May 29, 2024
The `param.grad is not None` check also fixes gradient reduction in the case of parameters not having acquired gradients (as parameters could become empty tensors in FSDP). Thanks to @ofivite for suggesting that `use_orig_params=True` could be the cause of the issue, which greatly helped with analysis. Signed-off-by: janEbert <janpublicebert@posteo.net>
|
Done thanks :) |
BoxiangW
pushed a commit
to BoxiangW/NeMo
that referenced
this pull request
Jun 5, 2024
The `param.grad is not None` check also fixes gradient reduction in the case of parameters not having acquired gradients (as parameters could become empty tensors in FSDP). Thanks to @ofivite for suggesting that `use_orig_params=True` could be the cause of the issue, which greatly helped with analysis. Signed-off-by: janEbert <janpublicebert@posteo.net> Signed-off-by: Boxiang Wang <boxiangw@nvidia.com>
janekl
pushed a commit
that referenced
this pull request
Jun 12, 2024
The `param.grad is not None` check also fixes gradient reduction in the case of parameters not having acquired gradients (as parameters could become empty tensors in FSDP). Thanks to @ofivite for suggesting that `use_orig_params=True` could be the cause of the issue, which greatly helped with analysis. Signed-off-by: janEbert <janpublicebert@posteo.net> Signed-off-by: Jan Lasek <janek.lasek@gmail.com>
rohitrango
pushed a commit
to rohitrango/NeMo
that referenced
this pull request
Jun 25, 2024
The `param.grad is not None` check also fixes gradient reduction in the case of parameters not having acquired gradients (as parameters could become empty tensors in FSDP). Thanks to @ofivite for suggesting that `use_orig_params=True` could be the cause of the issue, which greatly helped with analysis. Signed-off-by: janEbert <janpublicebert@posteo.net>
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.
As observed by some users in #8487, FSDP causes differences in loss. This is because gradients are not correctly calculated when
fsdp_use_orig_params=True. Currently, any non-FlatParameteris treated as being unsharded, when in reality they may be sharded whenuse_orig_params=True. This leads to a double reduction of sharded gradients. We thus simply adjust the parameters that are reduced manually. This requires use of a private variable in the FSDP module. Alternatively, the value ofself.kwargs['ignored_states']innlp_overrides.pycould be set as an attribute on the model to avoid this private variable access.Thanks to @ofivite for suggesting that
use_orig_params=Truecould be the cause of the issue, which greatly helped with analysis.What does this PR do ?
Fix FSDP gradient calculation when
fsdp_use_orig_params=True.Collection: nlp
Before your PR is "Ready for review"
Pre checks:
PR Type:
Who can review?
Anyone in the community is free to review the PR once the checks have passed.
Contributor guidelines contains specific people who can review PRs to various areas.
Additional Information