Target Layer Regularization for Continual Learning Using Cramer-Wold Distance - View it on GitHub