Use LRScheduler for torch >= 1.14 otherwise use _LRScheduler

[qmaruf]

What does this PR do?

Fixes #15587

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

No breaking changes.

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 🙃

[carmocca]

In general. If it was using torch.optim.lr_scheduler._LRScheduler, replace it with _TORCH_LRSCHEDULER

[qmaruf]

Getting this error after applying all suggested changes. @carmocca any suggestion?

ubuntu@ip-172-31-47-37:~/p/lightning{remote:feature-15587} $ python -m pytest tests/tests_pytorch/test_cli.py::test_lightning_cli_run                                       (lightning) 
================================================================================= test session starts ==================================================================================
platform linux -- Python 3.8.15, pytest-7.2.0, pluggy-1.0.0
rootdir: /home/ubuntu/pytorch-lightning/lightning, configfile: setup.cfg
plugins: forked-1.4.0, requests-mock-1.10.0, anyio-3.6.2, doctestplus-0.12.1, asyncio-0.20.2, rerunfailures-10.2, timeout-2.1.0, cov-4.0.0, hydra-core-1.2.0
asyncio: mode=strict
collected 1 item                                                                                                                                                                       

tests/tests_pytorch/test_cli.py F                                                                                                                                                [100%]

======================================================================================= FAILURES =======================================================================================
________________________________________________________________________________ test_lightning_cli_run ________________________________________________________________________________

cleandir = None

    def test_lightning_cli_run(cleandir):
        with mock.patch("sys.argv", ["any.py"]):
>           cli = LightningCLI(BoringModel, run=False)

/home/ubuntu/pytorch-lightning/lightning/tests/tests_pytorch/test_cli.py:817: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
/home/ubuntu/pytorch-lightning/lightning/src/pytorch_lightning/cli.py:345: in __init__
    _populate_registries(auto_registry)
/home/ubuntu/pytorch-lightning/lightning/src/pytorch_lightning/utilities/cli.py:128: in _populate_registries
    LR_SCHEDULER_REGISTRY.register_classes(torch.optim.lr_scheduler, LRScheduler, show_deprecation=False)
/home/ubuntu/pytorch-lightning/lightning/src/pytorch_lightning/utilities/cli.py:69: in register_classes
    for cls in self.get_members(module, base_cls):
/home/ubuntu/pytorch-lightning/lightning/src/pytorch_lightning/utilities/cli.py:77: in <genexpr>
    if issubclass(cls, base_cls) and cls != base_cls
/home/ubuntu/miniconda3/envs/lightning/lib/python3.8/abc.py:102: in __subclasscheck__
    return _abc_subclasscheck(cls, subclass)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

other = <class 'torch.optim.lr_scheduler.ChainedScheduler'>

    def _proto_hook(other):
        if not cls.__dict__.get('_is_protocol', False):
            return NotImplemented
    
        # First, perform various sanity checks.
        if not getattr(cls, '_is_runtime_protocol', False):
            if _allow_reckless_class_cheks():
                return NotImplemented
            raise TypeError("Instance and class checks can only be used with"
                            " @runtime_checkable protocols")
        if not _is_callable_members_only(cls):
            if _allow_reckless_class_cheks():
                return NotImplemented
>           raise TypeError("Protocols with non-method members"
                            " don't support issubclass()")
E           TypeError: Protocols with non-method members don't support issubclass()

/home/ubuntu/miniconda3/envs/lightning/lib/python3.8/typing.py:1081: TypeError
=============================================================================== short test summary info ================================================================================
FAILED tests/tests_pytorch/test_cli.py::test_lightning_cli_run - TypeError: Protocols with non-method members don't support issubclass()
============================================================================= 1 failed, 1 warning in 0.52s =============================================================================

[Borda]

@qmaruf could you pls check the conflicts? :)

[Borda]

@lantiga mind checking out this PR? 🐰