QLAB: Quadratic Loss Approximation-Based Optimal Learning Rate for Deep Learning

Open in new window