Check if optimizer supports closure

[tchaton]

What does this PR do?

Fixes #4974

Before submitting

• Was this discussed/approved via a Github issue? (no need for typos and docs improvements)
• 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? Otherwise, we ask you to create a separate PR for every change.
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?
• Did you verify new and existing tests pass locally with your changes?
• If you made a notable change (that affects users), did you update the CHANGELOG?

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; Bugfixes should be including in bug-fix release milestones (m.f.X) and features should be included in (m.X.b) releases.

Did you have fun?

Make sure you had fun coding 🙃

[pep8speaks]

Hello @tchaton! Thanks for updating this PR.

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-12-11 11:59:46 UTC

[awaelchli]

@tchaton I believe this is not correct. Every pytorch optimizer step requires a closure argument.
Code
https://pytorch.org/docs/stable/_modules/torch/optim/optimizer.html#Optimizer
Docs
https://pytorch.org/docs/stable/optim.html#torch.optim.Optimizer.step

[codecov]

Codecov Report

Merging #4981 (83faead) into master (4e6a871) will increase coverage by 0%.
The diff coverage is 100%.

@@          Coverage Diff           @@
##           master   #4981   +/-   ##
======================================
  Coverage      93%     93%           
======================================
  Files         133     133           
  Lines        9872    9876    +4     
======================================
+ Hits         9166    9170    +4     
  Misses        706     706           

[tchaton]

@tchaton I believe this is not correct. Every pytorch optimizer step requires a closure argument.
Code
https://pytorch.org/docs/stable/_modules/torch/optim/optimizer.html#Optimizer
Docs
https://pytorch.org/docs/stable/optim.html#torch.optim.Optimizer.step

I entirely agree, but we can't assume people will properly implement Optimizer.

With Bolt Lars Optimizer Wrapper.
Here is the code:

Inside of LightningOptimizer:

The error:

The PR introduces this fix:

And it works.

[rohitgr7]

wouldn't creating a simple custom optimizer by just overriding the step method will do the job instead of adding the whole bolts package as a requirement just for the LARSWrapper??

class CustomOptimizer(optim.Adam):
    def step(self):
        super().step()

[tchaton]

wouldn't creating a simple custom optimizer by just overriding the step method will do the job instead of adding the whole bolts package as a requirement just for the LARSWrapper??

class CustomOptimizer(optim.Adam):
    def step(self):
        super().step()

LARSWrapper was updated recently, it was missing a closure argument.