TensorFlow Decision Tree for Regression

I need a decision tree implemented using TensorFlow for a regression task. The data this model will be working with is mixed, meaning it includes both numerical and categorical data. The target is to achieve a moderate level of accuracy with this model. The model will need validation using SHAP explainer. Visualized using DTREEV. As well, forecasting using autoregressive model for time series forecasting. Can provide a functioning prototype, that utilized KNN.