-
Notifications
You must be signed in to change notification settings - Fork 21.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[UCC] Add pre & post processing for CPU collectives #89030
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/89030
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 69f2156: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
This pull request was exported from Phabricator. Differential Revision: D41291592 |
Summary: Pull Request resolved: pytorch#89030 The CPU block in `collective_post` was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed. Test Plan: Tested the reduce-scatter collective using PARAM Reviewed By: eastzone Differential Revision: D41291592 fbshipit-source-id: 9881745a04d950e7e169eb4d49aa818cb25488b0
db84c37
to
cb7f05b
Compare
This pull request was exported from Phabricator. Differential Revision: D41291592 |
1 similar comment
This pull request was exported from Phabricator. Differential Revision: D41291592 |
Summary: Pull Request resolved: pytorch#89030 The CPU block in `collective_post` was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed. Test Plan: Tested the reduce-scatter collective using PARAM Reviewed By: eastzone, kingchc Differential Revision: D41291592 fbshipit-source-id: 2f7b318be6cbcab433070e1c88da3a9f69a187f1
cb7f05b
to
c605a42
Compare
This pull request was exported from Phabricator. Differential Revision: D41291592 |
c605a42
to
a41e0c0
Compare
Summary: Pull Request resolved: pytorch#89030 The CPU block in `collective_post` was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed. Test Plan: Tested the reduce-scatter collective using PARAM Reviewed By: eastzone, kingchc Differential Revision: D41291592 fbshipit-source-id: 6f6e29e5585db43b87740ae35e1a7c24781f08e1
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM.
/easycla |
@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 |
Merge failedReason: 2 additional jobs have failed, first few of them are: trunk ,trunk / macos-12-py3-x86-64 / test (default, 2, 2, macos-12) Details for Dev Infra teamRaised by workflow job |
Summary: Pull Request resolved: pytorch#89030 The CPU block in `collective_post` was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed. Test Plan: Tested the reduce-scatter collective using PARAM Reviewed By: eastzone, kingchc Differential Revision: D41291592 fbshipit-source-id: c8e7a639d78039951d85e33509ac141934a8c837
a41e0c0
to
69f2156
Compare
This pull request was exported from Phabricator. Differential Revision: D41291592 |
@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 |
Summary: The CPU block in `collective_post` was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed. Test Plan: Tested the reduce-scatter collective using PARAM Reviewed By: eastzone Differential Revision: D41291592 Pull Request resolved: pytorch#89030 Approved by: https://github.com/kingchc, https://github.com/kwen2501
Summary: The CPU block in
collective_post
was missing pre & post processing. The reduce-scatter implementaion expects use of pre-processing callback to flatten the input tensors, however, the missing invocation meant grabage values were being passed.Test Plan: Tested the reduce-scatter collective using PARAM
Reviewed By: eastzone
Differential Revision: D41291592