xRBM is a python library that provides implementations of the family of Restericted Boltzmann Machine (RBM) models. It is still under development and its first vesrion will be released in Feb 2017.
Currently, xRBM includes implementations for the following:
- Restricted Boltzmann Machine (RBM)
- Conditional Restricted Boltzmann Machine (CRBM)
- CD-k Learning Algorithm
In future, I'm going to add implementations for the Factored CRBM (FCRBM) and the Persistent-CD learning algorithm.
You can install xRBM using pip:
pip install xrbm
Usage and Documentation
You can find the xRBM documentation here.
There are three tutorials in the Jupyter notebook format that guid you how to use the models in xRBM:
- Tutorial 1: Training an RBM on MNIST Dataset
- Tutorial 2: Training an RBM on MNIST Dataset
- Tutorial 3: Training a Conditional RBM on Timeseries Data
xRBM is provided under the MIT License. You can find the source code on the github repository.