Add implementation for irregular dimension selection for nested tensors. #88585

ani300 · 2022-11-07T16:21:45Z

Summary: This diff modifies the implementation of the select operator so slices of the irregular dimension can be selected (e.g. nt[:,0,:]).

Test Plan:
Added new unit tests to test that the new functions work as intended (see them in diff). To test,
buck test mode/dev-nosan //caffe2/test:nested

Differential Revision: D41083993

pytorch-bot · 2022-11-07T16:21:48Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/88585

📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓ Need help or want to give feedback on the CI? Visit our office hours

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit e0a79eb:
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

linux-foundation-easycla · 2022-11-07T16:21:49Z

The committers listed above are authorized under a signed CLA.

✅ login: ani300 / name: Antoni Viros (c51a6080d4fb4af72a928154f9911481a13436dc)

facebook-github-bot · 2022-11-07T16:22:32Z

This pull request was exported from Phabricator. Differential Revision: D41083993

Skylion007 · 2022-11-07T22:54:46Z

aten/src/ATen/native/nested/NestedTensorUtils.h

Suggested change

std::vector<Tensor> flat_tensors;

std::vector<Tensor> sizes;

std::vector<Tensor> flat_tensors;

flat_tensors.reserve(tensor_node.degree());

std::vector<Tensor> sizes;

sizes.reserve(tensor_node.degree());

facebook-github-bot · 2022-11-07T23:06:19Z

This pull request was exported from Phabricator. Differential Revision: D41083993

facebook-github-bot · 2022-11-08T15:04:45Z

This pull request was exported from Phabricator. Differential Revision: D41083993

cpuhrsch · 2022-11-08T18:41:13Z

aten/src/ATen/native/nested/NestedTensorMath.cpp

Since you only need this value of bounds checking, why not do this as part of the loop below?

facebook-github-bot · 2022-11-08T19:30:40Z

This pull request was exported from Phabricator. Differential Revision: D41083993

cpuhrsch · 2022-11-08T19:41:58Z

aten/src/ATen/native/nested/NestedTensorUtils.cpp

Just before we forget: This is now unused :)

Summary: Pull Request resolved: pytorch#88585 This diff modifies the implementation of the select operator so slices of the irregular dimension can be selected (e.g. nt[:,0,:]). Test Plan: Added new unit tests to test that the new function works as intended (see them in diff). To test, `buck test mode/dev-nosan //caffe2/test:nested` Reviewed By: cpuhrsch Differential Revision: D41083993 fbshipit-source-id: ec381764ab66878f2ff08dd1e23327136e77c78b

facebook-github-bot · 2022-11-08T19:59:49Z

This pull request was exported from Phabricator. Differential Revision: D41083993

ani300 · 2022-11-08T20:02:13Z

@pytorchbot merge -g

pytorchmergebot · 2022-11-08T20:04:04Z

Merge started

Your change will be merged once all checks on your PR pass since you used the green (-g) flag (ETA: 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

pytorchmergebot · 2022-11-08T20:44:21Z

Merge failed

Reason: The following mandatory check(s) failed (Rule superuser):

pull

Dig deeper by viewing the failures on hud

Details for Dev Infra team

Raised by workflow job

Summary: Pull Request resolved: pytorch#88585 This diff modifies the implementation of the select operator so slices of the irregular dimension can be selected (e.g. nt[:,0,:]). Test Plan: Added new unit tests to test that the new function works as intended (see them in diff). To test, `buck test mode/dev-nosan //caffe2/test:nested` Reviewed By: cpuhrsch Differential Revision: D41083993 fbshipit-source-id: 9666dc4e3ba7ef9e0edf3e3c655a1150eeaa5268

facebook-github-bot · 2022-11-08T21:03:49Z

This pull request was exported from Phabricator. Differential Revision: D41083993

ani300 · 2022-11-08T21:04:27Z

@pytorchbot merge -g

pytorchmergebot · 2022-11-08T21:07:49Z

Merge started

Your change will be merged once all checks on your PR pass since you used the green (-g) flag (ETA: 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

Implementation in #88585 should work for all dimensions. Removed unnecessary check that constrained select to dims 0 and 1 [ghstack-poisoned]

Implementation in #88585 should work for all dimensions. Removed unnecessary check that constrained select to dims 0 and 1 Pull Request resolved: #89150 Approved by: https://github.com/cpuhrsch

…rs. (pytorch#88585) Summary: This diff modifies the implementation of the select operator so slices of the irregular dimension can be selected (e.g. nt[:,0,:]). Test Plan: Added new unit tests to test that the new functions work as intended (see them in diff). To test, `buck test mode/dev-nosan //caffe2/test:nested` Differential Revision: D41083993 Pull Request resolved: pytorch#88585 Approved by: https://github.com/cpuhrsch

Implementation in pytorch#88585 should work for all dimensions. Removed unnecessary check that constrained select to dims 0 and 1 Pull Request resolved: pytorch#89150 Approved by: https://github.com/cpuhrsch

ani300 requested review from albanD, anjali411 and soulitzer as code owners November 7, 2022 16:21

pytorch-bot bot added the release notes: cpp release notes category label Nov 7, 2022

facebook-github-bot added the fb-exported label Nov 7, 2022

ani300 requested a review from cpuhrsch November 7, 2022 16:49

Skylion007 reviewed Nov 7, 2022

View reviewed changes

ani300 force-pushed the export-D41083993 branch from c7e7c5d to 5f12bd4 Compare November 7, 2022 23:06

ani300 force-pushed the export-D41083993 branch from 5f12bd4 to c51a608 Compare November 8, 2022 15:04

ani300 changed the title ~~Add implementation for foreach_add and add irregular dimension selection for nested tensors.~~ Add implementation for irregular dimension selection for nested tensors. Nov 8, 2022

cpuhrsch reviewed Nov 8, 2022

View reviewed changes

ani300 force-pushed the export-D41083993 branch from c51a608 to c516659 Compare November 8, 2022 19:30

ani300 requested a review from cpuhrsch November 8, 2022 19:32

cpuhrsch approved these changes Nov 8, 2022

View reviewed changes

cpuhrsch reviewed Nov 8, 2022

View reviewed changes

aten/src/ATen/native/nested/NestedTensorUtils.cpp Outdated

Copy link

Contributor

cpuhrsch Nov 8, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just before we forget: This is now unused :)

albanD removed their request for review November 8, 2022 19:43

ani300 force-pushed the export-D41083993 branch from c516659 to a0deeca Compare November 8, 2022 19:59

pytorch-bot bot added the ciflow/trunk Trigger trunk jobs on your pull request label Nov 8, 2022

ani300 force-pushed the export-D41083993 branch from a0deeca to e0a79eb Compare November 8, 2022 21:03

pytorchmergebot added the Merged label Nov 9, 2022

pytorchmergebot closed this in 9f58e02 Nov 9, 2022

mikaylagawarecki mentioned this pull request Nov 16, 2022

Removed unecessary check in select_nested #89150

Closed

mikaylagawarecki added a commit that referenced this pull request Nov 16, 2022

Update on "Removed unecessary check in select_nested"

2539515

Implementation in #88585 should work for all dimensions. Removed unnecessary check that constrained select to dims 0 and 1 [ghstack-poisoned]

mikaylagawarecki added a commit that referenced this pull request Nov 16, 2022

Update on "Removed unecessary check in select_nested"

4eb5a65

Implementation in #88585 should work for all dimensions. Removed unnecessary check that constrained select to dims 0 and 1 [ghstack-poisoned]

-    std::vector<Tensor> flat_tensors;
-    std::vector<Tensor> sizes;
+    std::vector<Tensor> flat_tensors;
+    flat_tensors.reserve(tensor_node.degree());
+    std::vector<Tensor> sizes;
+    sizes.reserve(tensor_node.degree());

Add implementation for irregular dimension selection for nested tensors. #88585

Add implementation for irregular dimension selection for nested tensors. #88585

Uh oh!

Conversation

ani300 commented Nov 7, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented Nov 7, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/88585

✅ No Failures

Uh oh!

linux-foundation-easycla bot commented Nov 7, 2022 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

facebook-github-bot commented Nov 7, 2022

Uh oh!

Skylion007 Nov 7, 2022

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Nov 7, 2022

Uh oh!

facebook-github-bot commented Nov 8, 2022

Uh oh!

cpuhrsch Nov 8, 2022

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Nov 8, 2022

Uh oh!

cpuhrsch Nov 8, 2022

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Nov 8, 2022

Uh oh!

ani300 commented Nov 8, 2022

Uh oh!

pytorchmergebot commented Nov 8, 2022

Merge started

Uh oh!

pytorchmergebot commented Nov 8, 2022

Merge failed

Uh oh!

facebook-github-bot commented Nov 8, 2022

Uh oh!

ani300 commented Nov 8, 2022

Uh oh!

pytorchmergebot commented Nov 8, 2022

Merge started

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

ani300 commented Nov 7, 2022 •

edited

Loading

pytorch-bot bot commented Nov 7, 2022 •

edited

Loading

linux-foundation-easycla bot commented Nov 7, 2022 •

edited

Loading