TL;DR: We present numerical evidence for the argument that if deep networks are trained using large (mini-)batches, they converge to sharp minimizers, and these minimizers have poor generalization properties.
Abstract: The stochastic gradient descent (SGD) method and its variants are algorithms of choice for many Deep Learning tasks. These methods operate in a small-batch regime wherein a fraction of the training data, say $32$--$512$ data points, is sampled to compute an approximation to the gradient. It has been observed in practice that when using a larger batch there is a degradation in the quality of the model, as measured by its ability to generalize. We investigate the cause for this generalization drop in the large-batch regime and present numerical evidence that supports the view that large-batch methods tend to converge to sharp minimizers of the training and testing functions---and as is well known, sharp minima lead to poorer generalization. In contrast, small-batch methods consistently converge to flat minimizers, and our experiments support a commonly held view that this is due to the inherent noise in the gradient estimation. We discuss several strategies to attempt to help large-batch methods eliminate this generalization gap.
Keywords: Deep learning, Optimization
Conflicts: northwestern.edu, intel.com, ibm.com
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 7 code implementations](https://www.catalyzex.com/paper/arxiv:1609.04836/code)