[CI] Fix Fast GPU tests on PR

[DN6]

What does this PR do?

Fast GPU on PR model tests are failing because of the size of the models (too small) and default split percents.

• Update the model init params and split percents so tests pass.
• Update disk offload test to try and force disk offloading even if the model is small.

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.

[sayakpaul]

Isn’t it a bit too much for CPUs?

[DN6]

Overall model size is still very small (2.6 MB). Adds 2 seconds to the current CPU tests.

[sayakpaul]

Think the CI should be trigged for changes to tests, too:

diffusers/.github/workflows/pr_tests.yml

Line 7 in 9a8e8db

paths: