Add dummy logger for internally disabling logging for some features

[SkafteNicki]

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 to update the docs?
• Did you write any new necessary tests?
• If you made a notable change (that affects users), did you update the CHANGELOG?

What does this PR do?

Fixes #1828 and #1830
Make sure that users code can run during tuner algorithms (learning rate finder and batch scaler) when they call loggers directly in their code.

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

[williamFalcon]

awesome!
in the case where i did this:

self.logger.experiment.add_image('images', grid, 0)

would the no-op still let any arbitrary function work like add_image since it's not defined? we basically need a way to allow any call here no?

I assume that's what this is doing?

class DummyExperiment(object):
    """ Dummy experiment """
    def nop(*args, **kw):
        pass

    def __getattr__(self, _):
        return self.nop

[williamFalcon]

Never mind lol... answered myself haha.

Well done!