fix tuner.scale_batch_size not finding batch size attribute when using datamodule

[awaelchli]

What does this PR do?

Fixes #5967

The tuner tries to detect if the model has an attribute for batch size, however before the trainer attaches the datamodule.
Hence the batch size attribute can't be found.

The newly added test fails on master.

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
• Check that target branch and milestone match!

Did you have fun?

Make sure you had fun coding 🙃

[codecov]

Codecov Report

Merging #5968 (840f74b) into master (34b733b) will decrease coverage by 3%.
The diff coverage is 100%.

@@           Coverage Diff           @@
##           master   #5968    +/-   ##
=======================================
- Coverage      94%     91%    -3%     
=======================================
  Files         161     161            
  Lines       11474   11368   -106     
=======================================
- Hits        10731   10339   -392     
- Misses        743    1029   +286     

[stale]

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://pytorch-lightning.readthedocs.io/en/latest/generated/CONTRIBUTING.html#pull-request or ask the assistance of a core contributor here or on Slack. Thank you for your contributions.

[edenlightning]

@awaelchli any plans to get this in?

[awaelchli]

Yes I try to get it done by the end of this week.

[carmocca]

Mildly related, why do we have this loop?

https://github.com/PyTorchLightning/pytorch-lightning/blob/34b733b35e02b13e01192a03ac4c3df89efa433a/pytorch_lightning/trainer/connectors/model_connector.py#L33

[awaelchli]

Mildly related, why do we have this loop?

legacy stuff, pre- accelerator refactor
the loop goes over the lightning module and the potentially wrapped lightning module.

this model connector class will be refactored-away.
In the end, the only thing we will ever want to set directly on the model is the trainer reference :)
and all others can be derived through it

[rohitgr7]

we have few tests related to tuner here: https://github.com/PyTorchLightning/pytorch-lightning/blob/master/tests/trainer/test_trainer_tricks.py

These should be added to tests/tuner too.

Else LGTM ✌️

[tchaton]

Nice fix !