[FSDP][2/N] Fix grad zero vs. None edge case
#87308
Closed
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
[ghstack-poisoned]
awgu
requested review from
mrshenli,
pritamdamania87,
zhaojuanmao,
rohan-varma,
H-Huang,
kwen2501 and
mingzhe09088
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/87308
Note: Links to docs will display an error until the docs builds have been completed. ✅ No Failures, 3 PendingAs of commit 08508fa: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: distributed (fsdp)
awgu
added a commit
that referenced
this pull request
Oct 19, 2022
ghstack-source-id: fd742a592feaea4b549d92ce5e025f4964cf2429 Pull Request resolved: #87308
[ghstack-poisoned]
awgu
added a commit
that referenced
this pull request
Oct 19, 2022
ghstack-source-id: 85614e658ab8a39020ed3908279ab1e80e0e4bc6 Pull Request resolved: #87308
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. This PR additionally 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
changed the title
None edge caseNone edge case
awgu
added a commit
that referenced
this pull request
Oct 19, 2022
ghstack-source-id: 85614e658ab8a39020ed3908279ab1e80e0e4bc6 Pull Request resolved: #87308
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. [ghstack-poisoned]
awgu
added a commit
that referenced
this pull request
Oct 19, 2022
ghstack-source-id: 83774374607bde9dd423c0c517c00eb84c8240c2 Pull Request resolved: #87308
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. [ghstack-poisoned]
awgu
added a commit
that referenced
this pull request
Oct 20, 2022
ghstack-source-id: b983aa8042152bec5f233812a4ebe05135df823b Pull Request resolved: #87308
zhaojuanmao
approved these changes
Oct 21, 2022
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. [ghstack-poisoned]
awgu
added a commit
that referenced
this pull request
Oct 21, 2022
ghstack-source-id: e392f6c6752548932f41ae921773de1b941a77b4 Pull Request resolved: #87308
|
@pytorchbot merge |
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 |
|
Hey @awgu. |
awgu
added a commit
to awgu/pytorch
that referenced
this pull request
Oct 22, 2022
ghstack-source-id: e392f6c6752548932f41ae921773de1b941a77b4 Pull Request resolved: pytorch#87308
This was referenced Oct 24, 2022
Closed
sgrigory
pushed a commit
to sgrigory/pytorch
that referenced
this pull request
Oct 28, 2022
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. Pull Request resolved: pytorch#87308 Approved by: https://github.com/zhaojuanmao
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Nov 5, 2022
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. Pull Request resolved: pytorch#87308 Approved by: https://github.com/zhaojuanmao
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Dec 10, 2022
Some original parameters corresponding to one `FlatParameter` may have `None` gradient while others do not. In that case, the `flat_param.grad` must be non-`None`. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a `_is_grad_none` mask over the parameters' gradients. - `_is_grad_none` is initialized to `False` for all. - `_is_grad_none[i]` is set to `True` when writing zeros in place of `None` when writing back the `i`th gradient. - `_is_grad_none[i]` is set to `False` via `_reset_is_grad_none()`, which should be called in the post-backward. See the docstring for details. - `_is_grad_none[i]` must be `False` in order to set `param.grad` to be a view into `flat_param.grad`. Pull Request resolved: pytorch#87308 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)NonegradientSome original parameters corresponding to one
FlatParametermay haveNonegradient while others do not. In that case, theflat_param.gradmust be non-None. However, FSDP should take care to expose the original parameters' gradients regardless. To achieve this, we track a_is_grad_nonemask over the parameters' gradients._is_grad_noneis initialized toFalsefor all._is_grad_none[i]is set toTruewhen writing zeros in place ofNonewhen writing back theith gradient._is_grad_none[i]is set toFalsevia_reset_is_grad_none(), which should be called in the post-backward. See the docstring for details._is_grad_none[i]must beFalsein order to setparam.gradto be a view intoflat_param.grad.