Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

Hi team! I have a PR to add flytecopilot support for Union types, should be ready for review. Thank you! <https://github.com/flyteorg/flyte/pull/6674> (context: <https://github.com/flyteorg/flyte/issues/6662>)

Hi <@U06FVMEJR6X>
Could you provide an example script that will originally trigger the error for me to test with? I tried following on master branch but it did not give the error.
Thank you!

```container_task = ContainerTask(
    name="square-container-task",
    cache_version="0.0.0",
    input_data_dir="/var/inputs",
    output_data_dir="/var/outputs",
    inputs=kwtypes(data_point=Union[int, float]),
    outputs=kwtypes(res=bool),
    image=ImageSpec(
        name="contianer_task_anomalies",
        base_image="python:3.10",
        registry="localhost:30000",
        builder="default",
        copy=["detect_anomalies.py"],
    ),
    command=[
        "python",
        "detect_anomalies.py",
        "{{.inputs.data_point}}",
        "/var/outputs",
    ],
)


@workflow
def detect_one_anomaly(data_point: int) -> bool:
    return container_task(data_point=data_point)```

Hm, that task looks like it should trigger the error. Maybe it needs to specify `io_strategy=IOStrategy.DOWNLOAD_EAGER` ? Also I believe the error is specific to remote executions, I'll add that detail to the original issue.