A machine learning model to predict molecular dipole moments

The model achieves highly accurate results compared to the current state of the art for predicting dipole moments at a fraction of the computational cost.