[Data] Update batch inference release tests

[bveeramani]

Why are these changes needed?

Other than minor cleanups, this PR:

• Makes batch inference release tests use autoscaling, to ensure Ray Data works well with autoscaling
• Removes the "raw images" variants, because there isn't much value to them in addition to the parquet tests, and people use parquet more often)
• Updates the node types, for consistency with the other release test compute

Related issue number

Checks

• I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
• I've run scripts/format.sh to lint the changes in this PR.
• I've included any doc changes needed for https://docs.ray.io/en/master/.
  • I've added any new APIs to the API Reference. For example, if I added a
    method in Tune, I've added it in doc/source/tune/api/ under the
    corresponding .rst file.
• I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
• Testing Strategy
  • Unit tests
  • Release tests
  • This PR is not tested :(

[bveeramani]

Not sure why this wasn't an issue before. Maybe because we override max_workers for the specific node type

[omatthew98]

lgtm.