Apply sharding based on priority & combine DistInfo and ExtraInfo#916
Closed
ejguan wants to merge 1 commit intometa-pytorch:mainfrom
Closed
Apply sharding based on priority & combine DistInfo and ExtraInfo#916ejguan wants to merge 1 commit intometa-pytorch:mainfrom
ejguan wants to merge 1 commit intometa-pytorch:mainfrom
Conversation
Contributor
|
@ejguan has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
Contributor
|
I did not know it was possible to land without GH stamp. |
Contributor
Author
|
@NivekT It's only possible for TorchData. PyTorch doesn't allow this anymore due to the GH first project |
This file contains hidden or 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
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
After pytorch/pytorch#88424 is landed, we are able to invoke
apply_shardingby sharding levels (distributed or multiprocessing). Then, we are able to give fine-control on sharding byReadingService.DistributedReadingService, we will only set sharding on the distributed levelPrototypeMPReadingService, we will set distributed sharding in the main process and set mp sharding in the worker processes. Previously, we set sharding in each worker process based on both distributed and mp information.worker_init_fndoesn't needDistInfoanymore. As, theDataPipehas been distributed sharded in the main process.DistInfoandExtraInfoforworker_reset_fnto synchronize the distributed seeds across distributed workers and set worker-local seeds based on both distributed and mp information.