JaggedTensor permute - less CPU ops #2786
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
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D70609204 |
|
This pull request was exported from Phabricator. Differential Revision: D70609204 |
che-sh
added a commit
to che-sh/torchrec
that referenced
this pull request
Mar 7, 2025
Summary: `JaggedTensor.permute` could be called with very large `indices` list (a few hundred items) - so calling python properties `self.keys()`, `self.variable_stride_per_key()` and `self.stride_per_key_per_rank()` in the loop over indices start to compound and take noticeable time **on CPU**. Differential Revision: D70609204
|
This pull request was exported from Phabricator. Differential Revision: D70609204 |
che-sh
added a commit
to che-sh/torchrec
that referenced
this pull request
Mar 7, 2025
Summary: Pull Request resolved: pytorch#2786 `JaggedTensor.permute` could be called with very large `indices` list (a few hundred items) - so calling python properties `self.keys()`, `self.variable_stride_per_key()` and `self.stride_per_key_per_rank()` in the loop over indices start to compound and take noticeable time **on CPU**. Differential Revision: D70609204
che-sh
force-pushed
the
export-D70609204
branch
from
a3d67d4 to
cc7219a
Compare
che-sh
added a commit
to che-sh/torchrec
that referenced
this pull request
Mar 18, 2025
Summary: `JaggedTensor.permute` could be called with very large `indices` list (a few hundred items) - so calling python properties `self.keys()`, `self.variable_stride_per_key()` and `self.stride_per_key_per_rank()` in the loop over indices start to compound and take noticeable time **on CPU**. Reviewed By: sarckk Differential Revision: D70609204
Summary: Pull Request resolved: pytorch#2786 `JaggedTensor.permute` could be called with very large `indices` list (a few hundred items) - so calling python properties `self.keys()`, `self.variable_stride_per_key()` and `self.stride_per_key_per_rank()` in the loop over indices start to compound and take noticeable time **on CPU**. Reviewed By: sarckk Differential Revision: D70609204
|
This pull request was exported from Phabricator. Differential Revision: D70609204 |
che-sh
force-pushed
the
export-D70609204
branch
from
cc7219a to
0d1e6d9
Compare
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.
Summary:
JaggedTensor.permutecould be called with very largeindiceslist (a few hundred items) - so calling python propertiesself.keys(),self.variable_stride_per_key()andself.stride_per_key_per_rank()in the loop over indices start to compound and take noticeable time on CPU.Differential Revision: D70609204