[Update] Print training log in rank0 #1296
Open
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
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Outdated suggestions cannot be applied.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.
Now the training log is printed in rank_last, it's not friendly to users if they change the amount of GPU node.
For example, a multi-node training job has master-0, worker-0, worker-1...
If we use 32 node(8 gpus each), we need to find the training log in worker-30.
Next time we use 64 node(8 gpus each), we need to find the training log in worker-62.
So print training log in rank0 is more friendly, we can just find the training log in master-0, no mater how many nodes we use for training.
Thanks.@jaredcasper @jon-barker