Calibrating Bayesian Generative Machine Learning for Bayesiamplification