XGBClassifier conversion leads to continuous model outputs

[Walid-Rahman2]

I have an xgboost classifier model that is trained to predict a binary target (0, 1). I followed the steps outlined in the notebooks, but the converted xgboost model predicts values on a continuous distribution between -6 and 10. The original base xgboost model does give prediction outputs of 0 and 1 only. Is there a decision boundary enforced by the volta model by which I could post-process the output, or does Volta not yet support classification? Thank you!

[VoltaML]

Hi @Walid-Rahman2. Thanks for trying our library. We only support regression for the time being. We'll have classification support soon.

[Walid-Rahman2]

Thank you, looking forward it 👍🏼