This repository is a collection of deep learning algorithms implemented using TensorFlow, intended as a command-line utility for quickly training and evaluating popular models as benchmarks. It provides implementations of convolutional networks, restricted Boltzmann machines, deep belief networks, stacked denoising autoencoders, and multilayer perceptrons. The package offers a scikit-learn-like interface with fit/predict methods and includes a pip package named yadlt for integration into Python code.