Remove use of local rank in trainers #926
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
Codecov ReportAttention: Patch coverage is
|
rayg1234
changed the title
lbluque
requested changes
Dec 17, 2024
| # TODO: dealing with local rank is dangerous | ||
| # T201111838 remove this and use CUDA_VISIBILE_DEVICES instead so trainers don't need to know about which devie to use | ||
| local_rank: int, | ||
| local_rank: int, # DEPRECATED, DO NOT USE |
There was a problem hiding this comment.
can we just remove this altogether and let if fail if used? or does it cause deeper issues if we do?
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Outdated suggestions cannot be applied.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.
Remove use local_rank to set cuda devices all together. Previously this was behind a feature flag but instead of maintaining the 2 versions, its easier to just consolidate since the local multiple-GPU runs are bugged right now and multiple ranks can be assigned on the same device. Each rank now will only see their local GPU (GPU-0) and not 8 devices.
After this there shouldn't be any code that tries to assign a cuda device that isn't "cuda" or "cuda:0".
To sanity check: