Adds Sampler Wrappers for custom samplers in distributed environment

[otaj]

What does this PR do?

Fixes #9883
Closes #12866, #12413

Adds a DistributedSamplerWrapper and UnrepeatedDistributedSamplerWrapper in order to wrap custom samplers in distributed environment.

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

None

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 🙃

cc @Borda @awaelchli @rohitgr7 @akihironitta @justusschock @ninginthecloud @otaj

[otaj]

could we clarify how close are we to the original code?

@Borda Where would you like us to clarify more? In documentation, in docstrings, in changelog?

[Borda]

could we clarify how close are we to the original code?

@Borda Where would you like us to clarify more? In documentation, in docstrings, in changelog?

probably in the same comment, you mention that it was inspired from...

[carmocca]

LGTM!

[otaj]

At this moment, I believe the PR is ready to go. We decided to move from Catalyst's implementation.

Currently, there is _DatasetSamplerWrapper which whole purpose is to wrap Sampler into a class with Dataset-like features (i.e. inheritance from Dataset and meaningful methods __getitem__ and __len__) and DistributedSamplerWrapper, which is very much just DistributedSampler which takes as a first argument Sampler instead of Dataset. Under the hood, we create _DatasetSamplerWrapper from the Sampler and just call the machinery of DistributedSampler

@carmocca , can you merge please if the final pass of review still looks good to you? Thanks!