moncarlo

moncarlo is a framework to build neural network models and inference with HMC neurons.

Currently this project is no longer under development. Feel free to look at the neurons/hmc.py code to understand how Hamiltonian Monte Carlo markov chains work. In those docs/code I've done my best to explain the important parts of how sampling from a Hamiltonian system works. To dive deeper, checkout Information Theory by the late David Mackay.

Features

Installation

Contribute

• Issue Tracker: github.com/jduyon/moncarlo/issues
• Source Code: github.com/jduyon/moncarlo

Support

License