[FSDP] Option to keep grads in lower precision #85062
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
Differential Revision: [D39529117](https://our.internmc.facebook.com/intern/diff/D39529117/) [ghstack-poisoned]
rohan-varma
requested review from
mrshenli,
pritamdamania87,
zhaojuanmao,
H-Huang,
awgu and
mingzhe09088
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/85062
Note: Links to docs will display an error until the docs builds have been completed. ✅ No Failures, 16 PendingAs of commit 336055f: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: distributed (fsdp)
This was referenced Sep 15, 2022
facebook-github-bot
added
the
oncall: distributed
rohan-varma
added a commit
that referenced
this pull request
Sep 15, 2022
Differential Revision: [D39529117](https://our.internmc.facebook.com/intern/diff/D39529117/) ghstack-source-id: 167439055 Pull Request resolved: #85062
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Outdated suggestions cannot be applied.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):
Rehash of a similar PR from a month ago that got stale. Adds a config to FSDP MP so that gradients can be kept in lower precision, to support optimizers such as AnyPrecisionOptimizer which would like to keep grads in bf16.
To do this, for sharded cases, we cannot simply omit the cast back to the full precision param dtype, otherwise when setting
p.grad = p._saved_grad_shardin finalize_params, autograd will throw an error indicating that the grad dtype should match the param dtype when it is being set.As a workaround, we re-cast after setting this. Although, this means that for cases that use gradient accumulation, p._saved_grad_shard will be of the reduced dtype because it is set to p.grad in
_prep_grad_for_backward. As a result, add a check + recast here as well.Differential Revision: D39529117