Skip to content

Fix passing Accelerator objects to the accelerator flag with devices=x & resolve training type plugin #10773

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.

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

Closed
wants to merge 7 commits into from

Conversation

kaushikb11
Copy link
Contributor

@kaushikb11 kaushikb11 commented Nov 26, 2021

What does this PR do?

Fixes #10592
Fixes #10775

Does your PR introduce any breaking changes? If yes, please list them.

Before submitting

  • Was this discussed/approved via a GitHub issue? (not for typos and docs)
  • Did you read the contributor guideline, Pull Request section?
  • 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 internal minor changes/refactorings)

PR review

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

  • 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

Did you have fun?

Make sure you had fun coding 🙃

cc @Borda @tchaton @akihironitta

@kaushikb11 kaushikb11 self-assigned this Nov 26, 2021
@kaushikb11 kaushikb11 added accelerator bug Something isn't working priority: 0 High priority task labels Nov 26, 2021
@kaushikb11 kaushikb11 marked this pull request as ready for review November 26, 2021 13:32
@kaushikb11 kaushikb11 changed the title Fix passing Accelerator objects to the accelerator flag with devices=x Fix passing Accelerator objects to the accelerator flag with devices=x & resolve training type plugin Nov 26, 2021
@kaushikb11 kaushikb11 changed the title Fix passing Accelerator objects to the accelerator flag with devices=x & resolve training type plugin Fix passing Accelerator objects to the accelerator flag with devices=x & resolve training type plugin Nov 26, 2021
@kaushikb11 kaushikb11 added this to the 1.5.x milestone Nov 26, 2021
@mergify mergify bot removed the has conflicts label Nov 27, 2021
- Fixed passing `Accelerator` objects to the accelerator flag with `devices=x` ([#10773](https://github.com/PyTorchLightning/pytorch-lightning/pull/10773))


- Resolve training type plugin when passed within `Accelerator` ([#10773](https://github.com/PyTorchLightning/pytorch-lightning/pull/10773))
Copy link
Contributor

Choose a reason for hiding this comment

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

I am confused by this fix. IMO it is not worth doing that considering the refactors that happen right now for 1.6. I could however see this being included in 1.5.x

Copy link
Contributor Author

Choose a reason for hiding this comment

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

@awaelchli But it would be required for 1.5.x, right?

Copy link
Contributor

Choose a reason for hiding this comment

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

Yes sure, it can be merged to master first, then be included in 1.5.x, but then tests that you add here become mostly outdated immediately after. That's all I'm saying.

Copy link
Contributor

Choose a reason for hiding this comment

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

In that case, let's just change the target branch to the bugfix release branch

Copy link
Contributor

@tchaton tchaton left a comment

Choose a reason for hiding this comment

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

LGTM !

]
assert not trainer.training_type_plugin.sync_batchnorm

trainer = Trainer(accelerator=CPUAccelerator(PrecisionPlugin(), DDPPlugin()), devices=2, sync_batchnorm=True)
Copy link
Contributor

Choose a reason for hiding this comment

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

Suggested change
trainer = Trainer(accelerator=CPUAccelerator(PrecisionPlugin(), DDPPlugin()), devices=2, sync_batchnorm=True)
trainer = Trainer(accelerator=CPUAccelerator(PrecisionPlugin(), DDPPlugin()), devices=2)

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Why remove it?

@mergify mergify bot added the ready PRs ready to be merged label Nov 30, 2021
@kaushikb11
Copy link
Contributor Author

Closing it. As the issue has been resolved after the AcceleratorConnector refactor

@kaushikb11 kaushikb11 closed this Mar 21, 2022
@Borda Borda deleted the fix/acc_type branch March 29, 2022 05:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
accelerator bug Something isn't working has conflicts priority: 0 High priority task ready PRs ready to be merged
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Resolve training type plugin when passed with Accelerator Support passing Accelerator objects to the accelerator flag with devices=x
6 participants