Disallow batch sampler with multiple IPU devices

[otaj]

What does this PR do?

When running on multiple IPUs, poptorch.DataLoader increases batch size, which is not compatible with batch_sampler argument at all. This was discovered during debugging of #13640

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

It is impossible to have a BatchSampler in while running on multiple IPUs. This never worked, however the user did not get an error message about it, but some different one, later on. Now the attempt to have explicit BatchSampler fails.

Note that users can still pass batch_size argument, which creates a default BatchSampler under the hood and this still works

Before submitting

• Was this discussed/approved via a GitHub issue? (not for typos and docs) (on Slack it was)
• 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 @justusschock @awaelchli @ninginthecloud @rohitgr7 @otaj @Borda @akihironitta

[codecov]

Codecov Report

Merging #13854 (7fdd6e7) into master (a90ef3b) will increase coverage by 27%.
The diff coverage is 82%.

@@            Coverage Diff            @@
##           master   #13854     +/-   ##
=========================================
+ Coverage      49%      76%    +27%     
=========================================
  Files         332      332             
  Lines       26000    26054     +54     
=========================================
+ Hits        12728    19812   +7084     
+ Misses      13272     6242   -7030