KLIFF is an interatomic potential fitting package that can be used to fit
physics-motivated (PM) potentials, as well as machine learning potentials such
as the neural network (NN) models.