Study on the Large Batch Size Training of Neural Networks Based on the Second Order Gradient

Abstract

Large batch size training in deep neural networks (DNNs) possesses a well-known 'generalization gap' that remarkably induces generalization performance degradation. However, it remains unclear how varying batch size affects the structure of a NN. Here, we combine theory with experiments to explore the evolution of the basic structural properties, including gradient, parameter update step length, and loss update step length of NNs under varying batch sizes. We provide new guidance to improve generalization, which is further verified by two designed methods involving discarding small-loss samples and scheduling batch size. A curvature-based learning rate (CBLR) algorithm is proposed to better fit the curvature variation, a sensitive factor affecting large batch size training, across layers in a NN. As an approximation of CBLR, the median-curvature LR (MCLR) algorithm is found to gain comparable performance to Layer-wise Adaptive Rate Scaling (LARS) algorithm. Our theoretical results and algorithm offer geometry-based explanations to the existing studies. Furthermore, we demonstrate that the layer wise LR algorithms, for example LARS, can be regarded as special instances of CBLR. Finally, we deduce a theoretical geometric picture of large batch size training, and show that all the network parameters tend to center on their related minima.

References in corpus (10)

• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
• Large Batch Training of Convolutional Networks
• An Empirical Model of Large-Batch Training
• AdaBatch: Adaptive Batch Sizes for Training Deep Neural Networks
• The Effects of Hyperparameters on SGD Training of Neural Networks
• Large batch size training of neural networks with adversarial training and second-order information
• SmoothOut: Smoothing Out Sharp Minima to Improve Generalization in Deep Learning
• Towards Theoretical Understanding of Large Batch Training in Stochastic Gradient Descent
• Proportionate gradient updates with PercentDelta
• Nonlinear Conjugate Gradients For Scaling Synchronous Distributed DNN Training