Density estimation likelihood-free inference
delfi is a Python toolbox for density estimation likelihood-free inference. It provides several algorithms for sequential neural posterior estimation (SNPE). Documentation including installation instructions and a tutorial on how to get started is available at http://www.mackelab.org/delfi/.
delfi‘s successor is sbi, a package for simulation-based inference based on PyTorch, see: https://github.com/mackelab/sbi.