fix mypy errors for loggers/wandb.py

[gautierdag]

What does this PR do?

This PR fixes mypy errors for the pytorch_lightning/loggers/wandb.py file.

Note, I had to loosen the typing in loggers.py to allow an Optional[str] return for both name and version functions of the Logger class.

There are a couple additional types that could be imported from wandb for stricter checks, for instance the ImageDataOrPathType type could be used to check valid image data. However, I resorted to keeping it Any for now.

Fixes pytorch_lightning/loggers/wandb.py from #13445

Does your PR introduce any breaking changes? If yes, please list them.

No

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 list all the breaking changes introduced by this pull request?
• Did you update the CHANGELOG? (not for typos, docs, test updates, or minor internal changes/refactors)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the 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 🙃

[otaj]

Hi, thanks for the PR, I left one comment in the code. It also seems one additional test gets broken by your change related to CLI (utilities/test_cli.py::test_wandb_logger_init_args). However, I'm not too familiar with CLI tests, would someone else mind taking a look at that test please? cc @carmocca as he helped land the particular failing test

[carmocca]

However, I'm not too familiar with CLI tests, would someone else mind taking a look at that test please? cc @carmocca as he helped land the particular failing test

I'm actually not sure about what's the problem. I see:

E               ValueError: Value "Namespace(class_path='pytorch_lightning.loggers.WandbLogger', init_args=Namespace(notes='wandb', save_dir='wandb'))" does not validate against any of the types in typing.Union[pytorch_lightning.loggers.logger.Logger, typing.Iterable[pytorch_lightning.loggers.logger.Logger], bool]:
E                 - Problem with given class_path "pytorch_lightning.loggers.WandbLogger":
E                   - Component not supported: Call(func=Attribute(value=Attribute(value=Name(id='self', ctx=Load()), attr='_wandb_init', ctx=Load()), attr='update', ctx=Load()), args=[], keywords=[keyword(arg=None, value=Name(id='kwargs', ctx=Load()))])

Perhaps you know @mauvilsa?

[gautierdag]

I'm actually not sure about what's the problem. I see:

E               ValueError: Value "Namespace(class_path='pytorch_lightning.loggers.WandbLogger', init_args=Namespace(notes='wandb', save_dir='wandb'))" does not validate against any of the types in typing.Union[pytorch_lightning.loggers.logger.Logger, typing.Iterable[pytorch_lightning.loggers.logger.Logger], bool]:
E                 - Problem with given class_path "pytorch_lightning.loggers.WandbLogger":
E                   - Component not supported: Call(func=Attribute(value=Attribute(value=Name(id='self', ctx=Load()), attr='_wandb_init', ctx=Load()), attr='update', ctx=Load()), args=[], keywords=[keyword(arg=None, value=Name(id='kwargs', ctx=Load()))])

Perhaps you know @mauvilsa?

Seems like the error is due to the "notes" key in the actual test 😕

@pytest.mark.skipif(not _WANDB_AVAILABLE, reason="wandb is required")
def test_wandb_logger_init_args():
    _test_logger_init_args("WandbLogger", {"save_dir": "wandb", "notes": "wandb"})

'Configuration check failed :: No action for destination key "notes" to check its value.'
 - Expected a List but got "Namespace(class_path='pytorch_lightning.loggers.WandbLogger', init_args=Namespace(name=None, save_dir='wandb', offline=False, id=None, anonymous=None, version=None, project=None, log_model=False, experiment=None, prefix='', agg_key_funcs=None, agg_default_func=None, notes='wandb'))"
  - Expected a <class 'bool'> but got "Namespace(class_path='pytorch_lightning.loggers.WandbLogger', init_args=Namespace(name=None, save_dir='wandb', offline=False, id=None, anonymous=None, version=None, project=None, log_model=False, experiment=None, prefix='', agg_key_funcs=None, agg_default_func=None, notes='wandb'))"

Changing "notes" to "name" like in the other logger tests fixes the issue. However I'm confused at how this test was even passing in the first place..

[codecov]

Codecov Report

Merging #13483 (7b2708e) into master (e3b29cb) will decrease coverage by 10%.
The diff coverage is 100%.

@@            Coverage Diff            @@
##           master   #13483     +/-   ##
=========================================
- Coverage      86%      76%    -10%     
=========================================
  Files         327      327             
  Lines       25485    25486      +1     
=========================================
- Hits        21893    19446   -2447     
- Misses       3592     6040   +2448