[sklearn.ColumnTransformer](https://github.com/scikit-learn/scikit-learn/blob/141123270a39c52a60e98017ca52795215dc2ce1/sklearn/compose/_column_transformer.py#L51-L78) uses nested lists of parameters to explain values of the `transformers` tuples. It renders nicely in the [generated documentation of sklearn](https://scikit-learn.org/stable/modules/generated/sklearn.compose.ColumnTransformer.html). We should be able to represent it as a nested list of lists too.