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

Flyte

Note, another input param in a different task defined in the same way except it’s of type pandas.DataFrame, works.
Why is that?
```def estimate(test_data: Optional[pd.DataFrame] = None, ...```

I'm guessing it's because `encoder` is of type that isn't understood by Flyte, whereas `pd.DataFrame` is a Flyte native type. But this should be supported. <@U0265RTUJ5B>, is this a bug?

this seems like a bug flyte should fallback on `FlytePickle` in cases where the type is not understood, unless this is a known limitation of the Union type, in which case the error message should be more informative. <@U044DJ094JG> can you share a minimally reproducible example?

Two things:
1. Optional parameters are not supported in tasks yet 
2. This looks like a bug in our support for unions. Samhita is right, this should be supported.

<@U01DYLVUNJE>,
The task’s input dataframe is either supplied by the caller or None if user doesn’t have a df as input.
I tried both, with or without Optional.
```def estimate(test_data: Optional[pd.DataFrame] = None, ...
def estimate(test_data: pd.DataFrame = None, ...```

In both cases, when None is the input, it will fail.

:ladybug: Create a new Flyte Bug issue: <https://github.com/flyteorg/flyte/issues/new?assignees=&amp;labels=bug%2Cuntriaged&amp;template=bug_report.yaml&amp;title=%5BBUG%5D+>

<@U044DJ094JG> would you mind filling out a bug report for the error? ideally it contains a copy-pastable example the repros the error

<@U01DYLVUNJE>, done. <https://github.com/flyteorg/flyte/issues/3319>