Differentiable Learning of Generalized Structured Matrices for Efficient Deep Neural Networks

arXiv:2310.18882v2 Announce Type: replace
Abstract: This paper investigates efficient deep neural networks (DNNs) to replace dense unstructured weight matrices with structured ones that possess desired properties. The challenge arises because the optimal weight matrix structure in popular neural network models is obscure in most cases and may vary from layer to layer even in the same network. Prior structured matrices proposed for efficient DNNs were mostly hand-crafted without a generalized framework to systematically learn them. To address this issue, we …

abstract arxiv cases challenge cs.ai cs.cv cs.lg differentiable eess.iv eess.sp generalized matrix network networks neural network neural networks paper popular type unstructured weight matrix