How to Train Your Neural Network: A Comparative Evaluation

Current state-of-the-art frameworks for large scale distributed deep learning and empirical results comparing their performance on large image and language training tasks are presented and algorithmic and implementation portions of each framework are discussed.