Fix tests related to DDP communication hooks

[akihironitta]

What does this PR do?

Part of #12373.

Fixes DDP communication hook tests that are not running in our current CI setup.

Description of the changes

• Removes DistributedDataParallel.get_ddp_logging_data() which only works with torch<1.9. In newer PyTorch, it's renamed to DistributedDataParallel._get_ddp_logging_data() -> Dict.
• Moves assertions to DDPStrategy.teardown because DistributedDataParallel is not available after trainer.fit() runs as it gets removed during its teardown (related to [bugfix] Reduce memory leaks #8490).

How to verify the changes

Note that the modified tests don't run in this PR. To verify the changes here, please either

• check out this GPU CI result running in Unpin CUDA docker image for GPU CI #12373 OR
• run tests locally with PyTorch 1.10+:

  # from the host
  git clone https://github.com/PyTorchLightning/pytorch-lightning.git
  cd pytorch-lightning
  docker pull pytorchlightning/pytorch_lightning:base-cuda-py3.7-torch1.11  # or torch1.10
  docker run --rm -it --gpus all --ipc host -v $(pwd):/work -w /work pytorchlightning/pytorch_lightning:base-cuda-py3.7-torch1.11  # or torch1.10
  
  # inside the container
  export PL_RUN_STANDALONE_TESTS=1
  python -m pytest tests/strategies/test_ddp_strategy_with_comm_hook.py::test_ddp_fp16_compress_comm_hook
  python -m pytest tests/strategies/test_ddp_strategy_with_comm_hook.py::test_ddp_sgd_comm_hook
  python -m pytest tests/strategies/test_ddp_strategy_with_comm_hook.py::test_ddp_fp16_compress_wrap_sgd_comm_hook
  python -m pytest tests/strategies/test_ddp_strategy_with_comm_hook.py::test_ddp_post_local_sgd_comm_hook

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?
• [n/a] Did you make sure to update the documentation with your changes? (if necessary)
• [n/a] 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?
• [n/a] 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 @akihironitta @justusschock @kaushikb11 @awaelchli @rohitgr7

[carmocca]

Nice!