dsgibbons

MLflow Pipelines look great, and I am excited to see where it will go. My main feature requests at this point would be: - Better support for unstructured datasets that...

> End to end tracking server configuration for MLflow with SQLite [scenario 2](https://mlflow.org/docs/latest/tracking.html#scenario-2-mlflow-on-localhost-with-sqlite). Could I please work on this item?

I've started working on scenario 2 @BenWilson2. I'll probably also work on > A full example of using child runs to track a hyperparameter optimization run (logging to scenario 2...

I can no longer work on this item. Please feel free to assign this to someone else @BenWilson2. I'm not sure how to remove my commit reference messages from this...

> Would be great if anyone else using a Mac could verify this behavior. Same happened to me on Mac today. Upgrading also fixed the issue for me.

> Thanks @dsgibbons, this helps. In order to help pinpoint the cause, could you share which editor and package manager are you using? > > PyCharm and poetry for me....

@mikaelho I'm using a 2015 MacBook Pro so no M1 here.

Failure case just needs new hash generating due to change to self._feature_names. Doesn't change rest of PR so feel free to review.

Figure layout changed should hopefully be fixed by https://github.com/shap/shap/pull/3200

Can you please add a minimal example in a formatted code block @PabloOlivaresUV?