interpretml / DiCE

Generate Diverse Counterfactual Explanations for any machine learning model.
https://interpretml.github.io/DiCE/
MIT License
1.3k stars 183 forks source link

DiCE for Custom Model Input #431

Open VarunRavi95 opened 4 months ago

VarunRavi95 commented 4 months ago

Hello, This is a great package!

However, I need some help, I have a Convolutional 1D Autoencoder for Anomaly Detection, where my forward method takes in two inputs separately for continuous and binary features. Is there anyway to generate counterfactuals for such a model, since right now, it says my forward method requires two inputs.