Skip to content

Fix wrong behavior of DDPStrategy option with simple GAN training using DDP #20936

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
wants to merge 15 commits into
base: master
Choose a base branch
from
Open

Fix wrong behavior of DDPStrategy option with simple GAN training using DDP #20936

wants to merge 15 commits into from
+165 −16

Conversation

samsara-ku
Copy link

@samsara-ku samsara-ku commented Jun 25, 2025 •
edited by SkafteNicki
Loading

What does this PR do?

Fixes #20866
Fixes #20328
Fixes #18740
Fixes #17212

This PR adds MultiModelDDPStrategy class and its simple execution example, for the multi-gpu training with GAN training.

Simply speaking:

* right way using DDP in pytorch.
# each model is divided and wrapped by DDP class
generator, discriminator = DDP(generator), DDP(discrimiantor)

* current pytorch-lightning GAN training.
# trainer contains two nn.Modules; generator and discriminator
DDP(trainer) -> trainer contains two nn.Modules; generator and discriminator

* proposed `MultiModelDDP` class can use this way.
# trainer also has two nn.Modules, generator and discriminator;
# this way allocate DDP class to the each model in the trainer, not the trainer.
MultiModelDDP(trainer)

  • Currently, pytorch lightning simple GAN training has has problem with DistributedDataParallel strategy. It tries to wrap pl.trainer, not the nn.Module models in the pl.trainer

  • Although we can activate find_unused_parameters=True options to avoid this issue but it is not right way; I think it is just a trick.

  • So the key idea to solve this issue is that we assign DistributedDataParallel to the each model in the pl.trainer, different from the previous strategy DDPStrategy.

  • I already tested with my GPUs to visaulize the result and tracked the gradients of model with each epoch; it works and you can see the visulized result in thie google drive link

Before submitting
  • Was this discussed/agreed via a GitHub issue? (not for typos and docs)
  • [v] Did you read the contributor guideline, Pull Request section?
  • [v] Did you make sure your PR does only one thing, instead of bundling different changes together?
  • Did you make sure to update the documentation with your changes? (if necessary)
  • Did you write any new necessary tests? (not for typos and docs)
  • Did you verify new and existing tests pass locally with your changes?
  • Did you list all the breaking changes introduced by this pull request?
  • Did you update the CHANGELOG? (not for typos, docs, test updates, or minor internal changes/refactors)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:

Reviewer checklist
  • Is this pull request ready for review? (if not, please submit in draft mode)
  • Check that all items from Before submitting are resolved
  • Make sure the title is self-explanatory and the description concisely explains the PR
  • Add labels and milestones (and optionally projects) to the PR so it can be classified

📚 Documentation preview 📚: https://pytorch-lightning--20936.org.readthedocs.build/en/20936/

@github-actions github-actions bot added the pl Generic label for PyTorch Lightning package label Jun 25, 2025
Borda
Borda reviewed Jun 25, 2025
View reviewed changes
Borda
Borda reviewed Jun 25, 2025
View reviewed changes
@stale Stale
Copy link

stale bot commented Jul 19, 2025

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. If you need further help see our docs: https://lightning.ai/docs/pytorch/latest/generated/CONTRIBUTING.html#pull-request or ask the assistance of a core contributor here or on Discord. Thank you for your contributions.

@stale stale bot added the won't fix This will not be worked on label Jul 19, 2025
@stale stale bot removed the won't fix This will not be worked on label Aug 5, 2025
Borda
Borda reviewed Aug 8, 2025
View reviewed changes
Borda
Borda reviewed Aug 8, 2025
View reviewed changes
Borda
Borda reviewed Aug 8, 2025
View reviewed changes
@Borda
Copy link
Member

Borda commented Aug 8, 2025

@SkafteNicki could you pls check too :)

Borda
Borda reviewed Aug 8, 2025
View reviewed changes
SkafteNicki
SkafteNicki reviewed Aug 9, 2025
View reviewed changes
@Borda
Copy link
Member

Borda commented Aug 11, 2025

@samsara-ku please mark also the comments as resolved to keep track what has left :)

SkafteNicki
SkafteNicki reviewed Aug 12, 2025
View reviewed changes
@Borda Borda requested a review from SkafteNicki August 12, 2025 17:13
SkafteNicki
SkafteNicki reviewed Aug 13, 2025
View reviewed changes
Copy link
Collaborator

@SkafteNicki SkafteNicki left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we are getting close with this PR, just a few suggestions

tests/tests_pytorch/strategies/test_multi_model_ddp.py
from lightning.pytorch.strategies.ddp import MultiModelDDPStrategy


def test_multi_model_ddp_setup_and_register_hooks():
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

please add one line docstring what the test does

tests/tests_pytorch/strategies/test_multi_model_ddp.py
)


def test_multi_model_ddp_register_hooks_cpu_noop():
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

please add one line docstring what the test does

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@Borda Borda Borda approved these changes

@SkafteNicki SkafteNicki SkafteNicki left review comments

@lantiga lantiga Awaiting requested review from lantiga lantiga is a code owner

@tchaton tchaton Awaiting requested review from tchaton tchaton is a code owner

@justusschock justusschock Awaiting requested review from justusschock justusschock is a code owner

@ethanwharris ethanwharris Awaiting requested review from ethanwharris ethanwharris is a code owner

Assignees
No one assigned
Labels
pl Generic label for PyTorch Lightning package
Projects
None yet
Milestone
No milestone
3 participants
@samsara-ku @Borda @SkafteNicki