[FSDP][optim_state_dict] Ignore the state check on rank that does not own the corresponding parameter #93318
Conversation
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
… own the corresponding parameter When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) [ghstack-poisoned]
fegin
requested review from
H-Huang,
awgu,
kwen2501,
mrshenli,
rohan-varma,
wanchaol and
zhaojuanmao
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/93318
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 1c9da19: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: distributed (fsdp)
fegin
added a commit
that referenced
this pull request
Jan 30, 2023
… own the corresponding parameter When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) ghstack-source-id: 178833186 Pull Request resolved: #93318
…at does not own the corresponding parameter" When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) [ghstack-poisoned]
fegin
added a commit
that referenced
this pull request
Jan 31, 2023
… own the corresponding parameter Pull Request resolved: #93318 When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. ghstack-source-id: 178911152 Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/)
rohan-varma
reviewed
Feb 1, 2023
Contributor
rohan-varma
left a comment
rohan-varma
left a comment
There was a problem hiding this comment.
LGTM overall, but could we add an example unittest that would've raised an error / correctness issue that this change would fix?
| local_state = ( | ||
| optim_state[name] | ||
| if name in optim_state | ||
| else torch.empty(max_numel, dtype=dtype, device=fsdp_state.compute_device) |
…at does not own the corresponding parameter" When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) [ghstack-poisoned]
…at does not own the corresponding parameter" When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) [ghstack-poisoned]
fegin
added a commit
that referenced
this pull request
Feb 1, 2023
… own the corresponding parameter Pull Request resolved: #93318 When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. ghstack-source-id: 179074159 Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/)
…at does not own the corresponding parameter" When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/) [ghstack-poisoned]
fegin
added a commit
that referenced
this pull request
Feb 2, 2023
… own the corresponding parameter Pull Request resolved: #93318 When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one. ghstack-source-id: 179122260 Differential Revision: [D42865237](https://our.internmc.facebook.com/intern/diff/D42865237/)
Contributor
Author
|
@pytorchbot merge |
|
This PR needs to be approved by an authorized maintainer before merge. |
rohan-varma
approved these changes
Feb 3, 2023
Contributor
rohan-varma
left a comment
rohan-varma
left a comment
There was a problem hiding this comment.
stamp to unblock but would be great if we can add the unittest or file an issue to do so.
Contributor
Author
Oh, this PR already has a UT for that. |
Contributor
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
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.
Stack from ghstack (oldest at bottom):
When a rank does not own a parameter (parameter.numel() == 0), its optim state is not valid and should not be checked against the current saved one.
Differential Revision: D42865237