Fix Pickling Issue with SWA and AdvancedProfiler

[justusschock]

What does this PR do?

Fixes #6522

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 🙃

I am not sure, why the AdvancedProfiler cannot be pickled. There seems to be an issue with a local function. But I was not able to find any. So this doesn't fix that but bypasses it!

[pep8speaks]

Hello @justusschock! Thanks for opening this PR.

• In the file pytorch_lightning/callbacks/stochastic_weight_avg.py:

Line 140:1: W293 blank line contains whitespace

Do see the Hitchhiker's guide to code style

[kaushikb11]

@justusschock Could we add a test for this, of Pickling not having an Issue?

[justusschock]

@kaushikb11 I'll try to come up with one, but not sure, how good it will be since this is something non-trivial to test

[tchaton]

Hey @justusschock,

I would say the bug should be resolved on the profiler side, not the callback side.

Best,
T.C

[carmocca]

The issue will be partly resolved by #6621

However, we still have the problem of not being able to pickle cProfiler.Profile

Can reproduce it with this test in test_profilers.py

def test_advanced_profiler_cprofile_deepcopy(tmpdir):
    """Checks for pickle issue reported in #6522"""
    model = BoringModel()
    trainer = Trainer(
        default_root_dir=tmpdir,
        fast_dev_run=True,
        profiler="advanced",
        stochastic_weight_avg=True,
    )
    trainer.fit(model)

edit: Closing in favor of fixes included in #6621