CARVIEW |
Navigation Menu
-
Notifications
You must be signed in to change notification settings - Fork 24.7k
Add wait counter for nccl abort #136067
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
Add wait counter for nccl abort #136067
Conversation
đź”— Helpful Linksđź§Ş See artifacts and rendered test results at hud.pytorch.org/pr/136067
Note: Links to docs will display an error until the docs builds have been completed. âś… No FailuresAs of commit 75445cc with merge base 48d18fb ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This pull request was exported from Phabricator. Differential Revision: D62675010 |
@atuljangra do you mind adding a brief comment above the code describing what the code does? |
@kwen2501 Added some more details int eh description. |
@atuljangra Thanks. Maybe I confused you :) What I meant was to copy the PR description into the code as a comment (so that readers know what that line of code is for). |
@kwen2501 Ah got it, will update. |
1717e5d
to
fa00385
Compare
This pull request was exported from Phabricator. Differential Revision: D62675010 |
fa00385
to
f660774
Compare
This pull request was exported from Phabricator. Differential Revision: D62675010 |
f660774
to
4d5f6c6
Compare
This pull request was exported from Phabricator. Differential Revision: D62675010 |
4d5f6c6
to
1fb266f
Compare
This pull request was exported from Phabricator. Differential Revision: D62675010 |
Summary: Pull Request resolved: pytorch#136067 Quite a few times, we see the NCCL PG abort taking too long. There's no easy way to measure this, so let's add a counter to measure this across the stack. Example job where this would have been useful: f641288746-TrainingApplication Test Plan: Unit tests https://www.internalfb.com/mlhub/pipelines/runs/mast/aps-atuljangra-8519d57d57?job_attempt=0&env=PRODUCTION Reviewed By: c-p-i-o Differential Revision: D62675010
1fb266f
to
75445cc
Compare
This pull request was exported from Phabricator. Differential Revision: D62675010 |
@pytorchbot merge -f 'Landed internally' (Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally) |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Summary: Quite a few times, we see the NCCL PG abort taking too long. There's no easy way to measure this, so let's add a counter to measure this across the stack. This will help us measure how much time we take the NCCL abort. Test Plan: Unit tests Reviewed By: c-p-i-o Differential Revision: D62675010 Pull Request resolved: pytorch#136067 Approved by: https://github.com/fduwjj
Summary:
Quite a few times, we see the NCCL PG abort taking too long. There's no easy way to measure this, so let's add a counter to measure this across the stack.
This will help us measure how much time we take the NCCL abort.
Test Plan:
Unit tests
Reviewed By: c-p-i-o
Differential Revision: D62675010
cc @XilunWu @H-Huang @awgu @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o