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

Flyte

image.png

Hey all, running through the Flyte school `00-intro` module and came across a peculiar error. It seems as though when i try to get the regression model we trained from the workflow, it is a `numpy.ndarray` instead of the expected `LogisticRegression` type. This causes errors later in the notebook when trying to use it to predict on a feature set. Any ideas on potential issues here?

Also, when i rerun the exact same code block, it seems to delete the element `o0` from the `execution_outputs`  . Is this expected behavior?

the printing does save what’s been converted.

so it should move from one section to the other, but it shouldn’t disappear

Screenshot 2023-09-22 at 10.59.36 AM.png

<@U05TDSZH309>, I did come across this error. When I used `joblib.load` to load the model, it worked.