Kedro is an open-sourced Python framework for creating maintainable and modular data science code.

Kedro

Hi friends, my team and I have developed an `MLFlowDataSet` handler because we wanted to use it as our model registry. It's just a thin wrapper around <https://www.mlflow.org/docs/latest/python_api/mlflow.sklearn.html> . Could this potentially be interesting for others as well and might be a worthwhile contribution?

think
```    def _save(self, data) -&gt; None:
        mlflow.set_tracking_uri(self.tracking_uri)
        model_info = mlflow.sklearn.log_model(
            sk_model=data,
            artifact_path=self.artifact_path,
            registered_model_name=self.registered_model_name,
        )```
and
```    def _load(self) -&gt; Any:
        [...]

        return mlflow.sklearn.load_model(mlflow_model_uri)```

How's it relate/differ from <https://pypi.org/project/kedro-mlflow/?|https://pypi.org/project/kedro-mlflow/?>

(Wonder if it might be a good addition to that plugin)

Very late, but agree that I'm not sure how it would differ from what Kedro-MLflow offers. From a previous look through that plugin, I think the datasets are implemented in a pretty top-notch way.