Reliable amortized variational inference with physics-based latent distribution correction

Code to partially reproduce results in "Reliable amortized variational inference with physics-based latent distribution correction".

Installation

Before starting installing the required packages in Julia, make sure you have matplotlib and seaborn installed in your Python environment since we depend on PyPlot.jl and Seaborn.jl for creating figures.

Run the command below to install the required packages.

julia -e 'Pkg.add("DrWatson.jl")'
julia --project -e 'using Pkg; Pkg.instantiate()'

The training dataset will download automatically into data/training-data/ directory upon running your first example.

Questions

Please contact alisk@rice.edu for questions.

Author

Ali Siahkoohi