[FSDP][1/N] Update summon_full_params(with_grads) None gradient
#87314
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
[ghstack-poisoned]
awgu
requested review from
H-Huang,
kwen2501,
mingzhe09088,
mrshenli,
pritamdamania87,
rohan-varma and
zhaojuanmao
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/87314
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 988a6ea: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: distributed (fsdp)
… gradient" This PR changes `summon_full_params(with_grads=True)`'s behavior to be such that if all ranks have `flat_param.grad = None`, then the original parameters will correctly have `orig_param.grad = None`. This is achieved with a preliminary all-reduce. Note that if a particular original parameter's gradient is `None` on all of the containing ranks, but not all ranks' `flat_param.grad = None`, then that particular gradient is still going to be set to zeros. This can be handled if desired in follow-up work. [ghstack-poisoned]
zhaojuanmao
approved these changes
Oct 21, 2022
| @torch.no_grad() | ||
| def unshard_grad(self): | ||
| """ | ||
| Unshards the handle's ``FlatParameter`` 's gradient. If all ranks have |
Contributor
There was a problem hiding this comment.
nit: maybe add a comment that 'unshard_grad' is not used in critical path, only used in summon_full_params(), as it calls all_reduce and may have performance impact
… gradient" This PR changes `summon_full_params(with_grads=True)`'s behavior to be such that if all ranks have `flat_param.grad = None`, then the original parameters will correctly have `orig_param.grad = None`. This is achieved with a preliminary all-reduce. Note that if a particular original parameter's gradient is `None` on all of the containing ranks, but not all ranks' `flat_param.grad = None`, then that particular gradient is still going to be set to zeros. This can be handled if desired in follow-up work. [ghstack-poisoned]
awgu
pushed a commit
to awgu/pytorch
that referenced
this pull request
Oct 21, 2022
ghstack-source-id: a0651cd Pull Request resolved: pytorch#87314
This was referenced Oct 24, 2022
Closed
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Nov 5, 2022
…ytorch#87314) This PR changes `summon_full_params(with_grads=True)`'s behavior to be such that if all ranks have `flat_param.grad = None`, then the original parameters will correctly have `orig_param.grad = None`. This is achieved with a preliminary all-reduce. Note that if a particular original parameter's gradient is `None` on all of the containing ranks, but not all ranks' `flat_param.grad = None`, then that particular gradient is still going to be set to zeros. This can be handled if desired in follow-up work. Pull Request resolved: pytorch#87314 Approved by: https://github.com/zhaojuanmao
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Dec 10, 2022
…ytorch#87314) This PR changes `summon_full_params(with_grads=True)`'s behavior to be such that if all ranks have `flat_param.grad = None`, then the original parameters will correctly have `orig_param.grad = None`. This is achieved with a preliminary all-reduce. Note that if a particular original parameter's gradient is `None` on all of the containing ranks, but not all ranks' `flat_param.grad = None`, then that particular gradient is still going to be set to zeros. This can be handled if desired in follow-up work. Pull Request resolved: pytorch#87314 Approved by: https://github.com/zhaojuanmao
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:
Noneedge case #87308 [FSDP][2/N] Fix grad zero vs.Noneedge casesummon_full_params(with_grads)Nonegradient #87314 [FSDP][1/N] Updatesummon_full_params(with_grads)NonegradientThis PR changes
summon_full_params(with_grads=True)'s behavior to be such that if all ranks haveflat_param.grad = None, then the original parameters will correctly haveorig_param.grad = None. This is achieved with a preliminary all-reduce. Note that if a particular original parameter's gradient isNoneon all of the containing ranks, but not all ranks'flat_param.grad = None, then that particular gradient is still going to be set to zeros. This can be handled if desired in follow-up work.