This script implements a simple Mixture Density Network in Pyro, a probabilistic programming language built on top of PyTorch. It uses the same example data as in Bishops's book (Bishop, C. M. (2013). Pattern Recognition and Machine Learning) and is based on the pytorch implementation of David Ha, https://github.com/hardmaru/pytorch_notebooks/blob/master/mixture_density_networks.ipynb
The code is located in a public repository hosted by Ascent Robotics (www.ascent.ai) at https://github.com/ascentai/mdn_pyro.