[core] Move ray.experimental.tf_utils back to rllib

[edoakes]

Why are these changes needed?

We have an experimental tensorflow utils library that was moved out of rllib and into core in 2019 (thread: #5698).

The way it's written makes Ray depend on RLlib (circular dependency). We could refactor this code to move the other related utils into Ray, but as far as I can tell, outside of rllib, this code was only used in an example from the documentation that was removed in 2022: #24727 (some of the code was missed, I'm removing it here).

We do not expose it in our documentation at all and it lives under experimental, so should be safe to remove.

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 :(

[edoakes]

@richardliaw PTAL and let me know if there's a concern I'm missing

[edoakes]

Not sure if this is right, lmk @sven1977 @simonsays1980

As far as I can tell we don't even have public docs for the tf_utils, so not sure if they're used