Leaping Forward and Learning: Overcoming the Challenges of TensorFlow - PROPRIUS