Fix Slow Tests

[DN6]

What does this PR do?

Switching to Torch 2.1 in the test runners resulted in a few failures on the slow tests.

This PR

1. Bumps up the precision / updates expected slices on a few of the failing tests
2. Updates the memory values in the SD v prediction test so that is passes. Torch 2.1 seems to provide more memory savings by default, so this value can be lowered.
3. Reintroduces pytorch_lightning as a test dependency. SD from_single_file checkpoints are serialised using pytorch_lightning modules, so these tests are breaking because we removed this dependency.
4. Breaks up T2IAdapters into individual tests for easier debugging

Fixes # (issue)

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline?
• Did you read our philosophy doc (important for complex PRs)?
• Was this discussed/approved via a Github issue or the forum? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[DN6]

@patrickvonplaten gentle ping.

[patrickvonplaten]

Thanks!