Python toolbox to create adversarial examples that fool neural networks - View it on GitHub