hanshupe

Support for taxonomies would be indeed an urgent feature, thanks for bringing it up.

I am very interested in this topic too - planning to use only the fine-tuning part and use the embeddings for semantic search. Any thoughts?

Did you compare it to supervised learning with fine-tuning?

Looks like that works: ``` 'sklearn.compose.TransformedTargetRegressor': { 'regressor': { 'sklearn.linear_model.LinearRegression': { 'normalize': [True, False], 'copy_X': [True] } }, 'func': {'numpy.log': None}, 'inverse_func': {'numpy.exp': None} }, ``` Is there a way...

Okay strange, don't know why it should be a transformer.

Thx!

I don't understand why that should not work. TPOT is optimizing the pipeline globally based on the defined scoring function and cross validation settings etc. So you can use the...

Can you reopen the question? I think it would be an important feature if it is not possible yet.

I use a custom configuration, which includes: ``` 'sklearn.feature_selection.SelectKBest': { 'k': range(3, 20), 'score_func': { 'sklearn.feature_selection.mutual_info_regression': None } ``` I would like to include additional score_func in the optimization process....

I guess you can just pickle the final pipeline object (tpot.fitted_pipeline_), which is a normal sklearn pipeline and can be used independently from tpot.