This is my implementation of a Reguly (2021) "Heterogeneous treatment effects in regression discontinuity designs" which you can find here. You can download the module causal_ml.py from here.