[export] optimize unflattener #115364
Closed
[export] optimize unflattener #115364
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
Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules). Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol. Differential Revision: [D51929775](https://our.internmc.facebook.com/intern/diff/D51929775/) [ghstack-poisoned]
suo
requested review from
avikchaudhuri,
gmagogsfm,
tugsbayasgalan and
zhxchen17
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/115364
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (1 Unrelated Failure)As of commit e42db18 with merge base 1224acc ( BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
suo
added a commit
that referenced
this pull request
Dec 7, 2023
Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules). Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol. Differential Revision: [D51929775](https://our.internmc.facebook.com/intern/diff/D51929775/) ghstack-source-id: 209551837 Pull Request resolved: #115364
Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules). Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol. Differential Revision: [D51929775](https://our.internmc.facebook.com/intern/diff/D51929775/) [ghstack-poisoned]
suo
added a commit
that referenced
this pull request
Dec 7, 2023
Pull Request resolved: #115364 Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules). Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol. ghstack-source-id: 209580532 @exported-using-ghexport Differential Revision: [D51929775](https://our.internmc.facebook.com/intern/diff/D51929775/)
zhxchen17
approved these changes
Dec 8, 2023
|
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
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 |
dmenig
pushed a commit
to dmenig/pytorch
that referenced
this pull request
Dec 21, 2023
Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules). Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol. Differential Revision: [D51929775](https://our.internmc.facebook.com/intern/diff/D51929775/) Pull Request resolved: pytorch#115364 Approved by: https://github.com/zhxchen17
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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):
Unflattening was slow on the APS FM model (which has thousands of nn.EmbeddingBag modules).
Quick glance at the profile shows 75% of time in unflattening was spent copying this node list, which is immutable and globally shared. So just passing it around as a tuple yields a 4x speedup lol.
Differential Revision: D51929775