Skip to content

Fix dataloaders are not reset when tuning the model#7566

Merged
awaelchli merged 5 commits intoLightning-AI:masterfrom
Lucklyric:fix/dl_not_rest_when_tune
May 24, 2021
Merged

Fix dataloaders are not reset when tuning the model#7566
awaelchli merged 5 commits intoLightning-AI:masterfrom
Lucklyric:fix/dl_not_rest_when_tune

Conversation

@Lucklyric
Copy link
Copy Markdown
Contributor

@Lucklyric Lucklyric commented May 16, 2021 •
edited
Loading

What does this PR do?

Fixes #7562

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 update the CHANGELOG? (not for typos, docs, test updates, or internal minor changes/refactorings)

PR review

Anyone in the community is free to review the PR once the tests have passed.
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 🙃

@codecov
Copy link
Copy Markdown

codecov Bot commented May 16, 2021 •
edited
Loading

Codecov Report

Merging #7566 (92d8f31) into master (e126649) will decrease coverage by 4%.
The diff coverage is 100%.

@@           Coverage Diff           @@
##           master   #7566    +/-   ##
=======================================
- Coverage      92%     88%    -4%     
=======================================
  Files         196     198     +2     
  Lines       12835   12837     +2     
=======================================
- Hits        11854   11291   -563     
- Misses        981    1546   +565

@Lucklyric Lucklyric marked this pull request as ready for review May 16, 2021 06:58
@Lucklyric Lucklyric changed the title Fix dataloaders are not rest when tuning the model Fix dataloaders are not reset when tuning the model May 17, 2021
carmocca
carmocca suggested changes May 17, 2021
View reviewed changes
Copy link
Copy Markdown
Contributor

@carmocca carmocca left a comment

Choose a reason for hiding this comment

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

Can you add a test that fails in current master and is fixed by the changes in this PR?

Comment thread pytorch_lightning/trainer/training_loop.py Outdated
Comment thread pytorch_lightning/trainer/training_loop.py Outdated
@carmocca carmocca added bug Something isn't working tuner labels May 18, 2021
@carmocca carmocca added this to the v1.3.x milestone May 18, 2021
@Lucklyric Lucklyric force-pushed the fix/dl_not_rest_when_tune branch from 4f8b5ab to d5c4c3f Compare May 18, 2021 18:59
@Lucklyric Lucklyric requested a review from SkafteNicki as a code owner May 18, 2021 18:59
tchaton
tchaton approved these changes May 21, 2021
View reviewed changes
Copy link
Copy Markdown
Contributor

@tchaton tchaton left a comment •
edited
Loading

Choose a reason for hiding this comment

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

LGTM ! Good catch ! Just need the tests.

carmocca
carmocca approved these changes May 21, 2021
View reviewed changes
Copy link
Copy Markdown
Contributor

@carmocca carmocca left a comment

Choose a reason for hiding this comment

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

Added the test that fails in current master myself. LGTM

carmocca
carmocca suggested changes May 21, 2021
View reviewed changes
Copy link
Copy Markdown
Contributor

@carmocca carmocca left a comment

Choose a reason for hiding this comment

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

Just noticed I'm not able to push changes to your fork's branch. I'm guessing you have disabled this:

https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests/working-with-forks/allowing-changes-to-a-pull-request-branch-created-from-a-fork

Can you enable it? Alternatively, you can cherry-pick this commit d88b88c

@Lucklyric
Copy link
Copy Markdown
Contributor Author

Lucklyric commented May 21, 2021

Dear @carmocca , it seems the option is already enabled. NVM, I have cherry-picked your revision. Thank you so much.

@carmocca carmocca added the ready to be merged PRs ready to be merged label May 21, 2021
awaelchli
awaelchli reviewed May 21, 2021
View reviewed changes
Comment thread pytorch_lightning/tuner/batch_size_scaling.py Outdated
@carmocca carmocca removed the ready to be merged PRs ready to be merged label May 21, 2021
ethanwharris
ethanwharris approved these changes May 24, 2021
View reviewed changes
Copy link
Copy Markdown
Member

@ethanwharris ethanwharris left a comment

Choose a reason for hiding this comment

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

LGTM 😃

@ethanwharris ethanwharris added the ready to be merged PRs ready to be merged label May 24, 2021
@awaelchli awaelchli merged commit 0c958c5 into Lightning-AI:master May 24, 2021
awaelchli added a commit that referenced this pull request May 25, 2021
@awaelchli @carmocca
Fix dataloaders are not reset when tuning the model (#7566)
d0a29b5 
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
@Borda Borda mentioned this pull request May 25, 2021
Closed
Borda pushed a commit that referenced this pull request May 26, 2021
@awaelchli @carmocca @Borda
Fix dataloaders are not reset when tuning the model (#7566)
e58af13 
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
@Borda Borda mentioned this pull request May 26, 2021
Merged
11 tasks
lexierule pushed a commit that referenced this pull request May 26, 2021
@awaelchli @carmocca @lexierule
Fix dataloaders are not reset when tuning the model (#7566)
f11c754 
Co-authored-by: Carlos Mocholi <carlossmocholi@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ethanwharris ethanwharris ethanwharris approved these changes

@tchaton tchaton tchaton approved these changes

@Borda Borda Awaiting requested review from Borda

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

@kaushikb11 kaushikb11 Awaiting requested review from kaushikb11

@SeanNaren SeanNaren Awaiting requested review from SeanNaren

@williamFalcon williamFalcon Awaiting requested review from williamFalcon

@SkafteNicki SkafteNicki Awaiting requested review from SkafteNicki

+2 more reviewers

@carmocca carmocca carmocca approved these changes

@awaelchli awaelchli awaelchli approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

bug Something isn't working ready to be merged PRs ready to be merged tuner

Projects

None yet

Milestone

pl:1.3.x

Development

Successfully merging this pull request may close these issues.

stopped calling datamodule.train_dataloader() when tuning the batch_size using pl.trainer.tune()

5 participants

@Lucklyric @awaelchli @ethanwharris @carmocca @tchaton