Python library for unsupervised training of Deep Restricted Bolztmann Machines using two different approaches. To run the MNIST experiment for a single RBM just type: python3 run.py from src folder with additional command line args specified as --argument value, for example: python3 run.py --epochs 50 --hiddenUnits 500 To run the MNIST experiment for a DBN specify the sizes of consecutive layers in runDBN.py and then: python3 runDBN.py The code is MIT licensed.