update torchmetrics to catch bug fix

[jpcbertoldo]

Description

I had issues to compute AUROC on a CPU in a use case where I didn't have enough memory in the GPU.

I opened an issue in tochmetrics: Lightning-AI/torchmetrics#1323

and it has been fixed: Lightning-AI/torchmetrics#1333

Could we upgrade torchmetrics to catch this bug fix?

Version 0.10.3 is not released yet but should come next week.

For future ref: https://github.com/Lightning-AI/metrics/tags

Disclaimer: I have not tried to run the pre-merge tests locally because apparently there are issues with them (#704)

Changes

• Bug fix (non-breaking change which fixes an issue)
• Refactor (non-breaking change which refactors the code base)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Checklist

• My code follows the pre-commit style and check guidelines of this project.
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing tests pass locally with my changes

[jpcbertoldo]

It's the good old coverage report causing problems : ) so I guess there's nothing breaking because of the version changing.

Although, the current requrirement is

torchmetrics>=0.9.1,<=0.9.3

Is there as specific reason for the <=0.9.3 or is it just for precaution?

[samet-akcay]

The tests are flaky. We still need to investigate. We could merge this one though. Thanks a lot!