Make shuffling optional in DistributedSampler #22479
Closed
Conversation
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
pytorchbot
added
the
module: dataloader
soumith
suggested changes
Jul 3, 2019
torch/utils/data/distributed.py
Outdated
Contributor
There was a problem hiding this comment.
break this out into a separate if/else block, instead inlining if/else inside indices = (
Contributor
|
I'm also really confused. If you dont need random shuffling, you dont need a DistributedSampler at all, right? |
Summary: Pull Request resolved: #22479 In some cases, for example, when we training on CTR data, we would like to start training from old samples and finish on new recent samples. This diff add the option to disable the shuffling in DistributedSampler to accommodate this use case. Differential Revision: D16100388 fbshipit-source-id: e0d31328920a4aa2315820683242cc44959c5d82
Contributor
|
okay, nevermind, this actually makes sense to effectively do sharding without shuffling. It took me a second to wrap my head around this |
soumith
approved these changes
Jul 6, 2019
Contributor
|
This pull request has been merged in 7730346. |
Collaborator
|
test? |
Contributor
|
my bad on the code review. i should've requested a test |
This was referenced Jul 12, 2019
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 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Outdated suggestions cannot be applied.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.
Summary:
In some cases, for example, when we training on CTR data, we would like to start training from old samples and finish on new recent samples.
This diff add the option to disable the shuffling in DistributedSampler to accommodate this use case.
Differential Revision: D16100388