This is an implementation of graphical models inference (Variable Elimination) using PyTorch for implementing factors (conditional probabilities and priors). Those can be implemented as tables (the typical discrete graphical model implementation) but also as arbitrary PyTorch modules, the case of most interest being of course neural networks.
The reason we are calling it "probabilistic programs" even though we only have discrete graphical models so far is that factors and variables are implemented in a manner general enough for probabilistic programming symbolic implementations be introduced down the road.
This project is for researchers interested in exploring the connection between probabilistic and neural network inference. It is a library that provides the utilities for mixing neural networks and graphical models while offering GPU use and stochastic gradient descent training.
This repository contains:
- source code, including an experiments directory
- test code (to be run with PyTest)
environment.ymlfor importing a Conda environment
In src/experiments one can find src/experiments/simple_mnist.py, which shows how to use a "graphical model" with a single factor,
implemented by a convolutional neural network, to learn how to recognize MNIST digits.
In src/experiments/mnist_pairs_semi_supervised.py there is code attempting to train a MNIST recognizer from pairs of images of successive digits only. It does not yet work but the script offers many options for running simplified versions of the problem, many of which do work.
Tests in test are split into quick_tests and slow_tests.
The former include tests of basic data structure implementations while the latter includes learning sessions with stochastic gradient descent and take several minutes.