Online Normalization for Training Neural Networks
Abstract
Online Normalization is a new technique for normalizing the hidden activations of a neural network. Like Batch Normalization, it normalizes the sample dimension. While Online Normalization does not use batches, it is as accurate as Batch Normalization. We resolve a theoretical limitation of Batch Normalization by introducing an unbiased technique for computing the gradient of normalized activations. Online Normalization works with automatic differentiation by adding statistical normalization as a primitive. This technique can be used in cases not covered by some other normalizers, such as recurrent networks, fully connected networks, and networks with activation memory requirements prohibitive for batching. We show its applications to image classification, image segmentation, and language modeling. We present formal proofs and experimental results on ImageNet, CIFAR, and PTB datasets.
Cite
@article{arxiv.1905.05894,
title = {Online Normalization for Training Neural Networks},
author = {Vitaliy Chiley and Ilya Sharapov and Atli Kosson and Urs Koster and Ryan Reece and Sofia Samaniego de la Fuente and Vishal Subbiah and Michael James},
journal= {arXiv preprint arXiv:1905.05894},
year = {2019}
}
Comments
Published at the Conference on Neural Information Processing Systems (NeurIPS 2019), Vancouver, Canada. Code: https://github.com/Cerebras/online-normalization