Fix ray remote options

[pcmoritz]

This PR fixes the prefect_ray.remote_options feature. Without this fix, it is possible to get an error of the form

AssertionError: The @ray.remote decorator must be applied either with no arguments and no parentheses, for example '@ray.remote', or it must be applied using some of the arguments in the list ['max_calls', 'max_retries', 'num_cpus', 'num_returns', 'object_store_memory', 'retry_exceptions', 'concurrency_groups', 'lifetime', 'max_concurrency', 'max_restarts', 'max_task_retries', 'max_pending_calls', 'namespace', 'get_if_exists', 'accelerator_type', 'memory', 'name', 'num_gpus', 'placement_group', 'placement_group_bundle_index', 'placement_group_capture_child_tasks', 'resources', 'runtime_env', 'scheduling_strategy', '_metadata'], for example '@ray.remote(num_returns=2, resources={"CustomResource": 1})'.

Closes #

Example

Checklist

• This pull request references any related issue by including "Closes #<ISSUE_NUMBER>"
  • If no issue exists and your change is not a small fix, please create an issue first.
• This pull request includes tests or only affects documentation.
• Summarized PR's changes in CHANGELOG.md

[ahuang11]

Curious when does the error happen? Would it be possible to replicate it in a test?

Also, to fix static analysis: pre-commit install && pre-commit run --all

[pcmoritz]

This actually happens if you just run the example in the README:

In [1]: from prefect import flow, task
   ...: from prefect_ray.task_runners import RayTaskRunner
   ...: from prefect_ray.context import remote_options
   ...: 
   ...: @task
   ...: def process(x):
   ...:     return x + 1
   ...: 
   ...: 
   ...: @flow(task_runner=RayTaskRunner())
   ...: def my_flow():
   ...:     # equivalent to setting @ray.remote(num_cpus=4, num_gpus=2)
   ...:     with remote_options(num_cpus=4, num_gpus=2):
   ...:         process.submit(42)
   ...: 

In [2]: my_flow()

Happy to add a test if you point me to where it should be added.

[pcmoritz]

I guess adding it in https://github.com/PrefectHQ/prefect-ray/blob/main/tests/test_task_runners.py should be fine? I'll give that a shot :)

[ahuang11]

Thanks! Can you add it under test_task_runners:
https://github.com/PrefectHQ/prefect-ray/blob/main/tests/test_task_runners.py#L232

It can literally be that example.

[ahuang11]

On second thought, maybe we should lower the number of CPUs and remove the number of GPUs requested in the test.

[pcmoritz]

Yes, happy to do that :)

I can also use the config that made me run into this -- max_calls, it won't require changes to the CPUs or GPUs :)

[pcmoritz]

Btw, is it possible to do a new release that includes this fix? Or if that's too disruptive, cut a release candidate with it so it can be installed with --pre? I'd like to get this to somebody ASAP :)

[ahuang11]

Yes I think we might be able to release tomorrow to fix this.

[desertaxle]

LGTM!