fix: ray module not found handling

[clumsy]

TorchX has been handling ModuleNotFoundError gracefully for a while now, e.g. for SageMaker when running torchx runopts we get:

...
            (remote jobs) the image repository to use when pushing patched images, must have push access. Ex: example.com/your/container
        quiet=QUIET (bool, False)
            whether to suppress verbose output for image building. Defaults to ``False``.

aws_sagemaker: No module named 'sagemaker'

gcp_batch:
    usage:
        [project=PROJECT],[location=LOCATION]
...

But for ray we get an exception after which we won't get next runopts:

gcp_batch:
    usage:
        [project=PROJECT],[location=LOCATION]

    optional arguments:
        project=PROJECT (str, None)
            Name of the GCP project. Defaults to the configured GCP project in the environment
        location=LOCATION (str, us-central1)
            Name of the location to schedule the job in. Defaults to us-central1

Traceback (most recent call last):
  File "/usr/local/bin/torchx", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.10/dist-packages/torchx/cli/main.py", line 118, in main
    run_main(get_sub_cmds(), argv)
  File "/usr/local/lib/python3.10/dist-packages/torchx/cli/main.py", line 114, in run_main
    args.func(args)
  File "/usr/local/lib/python3.10/dist-packages/torchx/cli/cmd_runopts.py", line 36, in run
    opts = runner.scheduler_run_opts(scheduler)
  File "/usr/local/lib/python3.10/dist-packages/torchx/runner/api.py", line 473, in scheduler_run_opts
    return self._scheduler(scheduler).run_opts()
  File "/usr/local/lib/python3.10/dist-packages/torchx/runner/api.py", line 718, in _scheduler
    sched = factory(self._name, **self._scheduler_params)
  File "/usr/local/lib/python3.10/dist-packages/torchx/schedulers/__init__.py", line 39, in run
    module = importlib.import_module(path)
  File "/usr/lib/python3.10/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1050, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1027, in _find_and_load
  File "<frozen importlib._bootstrap>", line 1006, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 688, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 883, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "/usr/local/lib/python3.10/dist-packages/torchx/schedulers/ray_scheduler.py", line 448, in <module>
    session_name: str, ray_client: Optional[JobSubmissionClient] = None, **kwargs: Any
NameError: name 'JobSubmissionClient' is not defined

That's because ray_scheduler has custom ModuleNotFoundException handling - perhaps for historic reasons.

Test plan:
[x] existing test must pass

[clumsy]

Please check this fix if you can, @d4l3k, @kiukchung, @tonykao8080

[andywag]

lgtm

[facebook-github-bot]

@andywag has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.