[Train] Immediately fail if application errors on any worker #28314
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
Signed-off-by: Amog Kamsetty <[email protected]>
Signed-off-by: Amog Kamsetty <[email protected]>
krfricke
added
the
@author-action-required
…n-immediate-error Signed-off-by: Amog Kamsetty <[email protected]>
Signed-off-by: Amog Kamsetty <[email protected]>
amogkam
removed
the
@author-action-required
Signed-off-by: Amog Kamsetty <[email protected]>
Signed-off-by: Amog Kamsetty <[email protected]>
|
@krfricke can you review again please? |
Signed-off-by: Amog Kamsetty <[email protected]>
Signed-off-by: Amog Kamsetty <[email protected]>
…n-immediate-error
7 tasks
amogkam
added a commit
that referenced
this pull request
Nov 2, 2022
Signed-off-by: Amog Kamsetty [email protected] Follow up to #28314 #28314 did not cover all the cases. In particular, if one worker fails, but the other workers are hanging, then our shutdown logic will also hang since it's waiting for the actors to finish running their methods. Instead, we want to force shutdown all workers regardless of if they have finished their method or not. This PR also adds an e2e integration test.
WeichenXu123
pushed a commit
to WeichenXu123/ray
that referenced
this pull request
Dec 19, 2022
Signed-off-by: Amog Kamsetty [email protected] Follow up to ray-project#28314 ray-project#28314 did not cover all the cases. In particular, if one worker fails, but the other workers are hanging, then our shutdown logic will also hang since it's waiting for the actors to finish running their methods. Instead, we want to force shutdown all workers regardless of if they have finished their method or not. This PR also adds an e2e integration test. Signed-off-by: Weichen Xu <[email protected]>
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.
Signed-off-by: Amog Kamsetty [email protected]
Resolves https://discuss.ray.io/t/torch-trainer-gets-stuck/7447
Previously with Ray Train, if 1 worker raises an Exception in the application code, this Exception is not raised in the driver until all workers have finished executing the training function. This can lead to hanging behaviors where other workers are waiting on a collective/
torch.distributed.barrier()call for example.With this PR, if any worker fails in the application code, the entire training job is immediately terminated and the exception is raised on the driver.
Why are these changes needed?
Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.