Spectral Tensor Train Parameterization of Deep Learning Layers
Abstract
We study low-rank parameterizations of weight matrices with embedded spectral properties in the Deep Learning context. The low-rank property leads to parameter efficiency and permits taking computational shortcuts when computing mappings. Spectral properties are often subject to constraints in optimization problems, leading to better models and stability of optimization. We start by looking at the compact SVD parameterization of weight matrices and identifying redundancy sources in the parameterization. We further apply the Tensor Train (TT) decomposition to the compact SVD components, and propose a non-redundant differentiable parameterization of fixed TT-rank tensor manifolds, termed the Spectral Tensor Train Parameterization (STTP). We demonstrate the effects of neural network compression in the image classification setting and both compression and improved training stability in the generative adversarial training setting.
Cite
@article{arxiv.2103.04217,
title = {Spectral Tensor Train Parameterization of Deep Learning Layers},
author = {Anton Obukhov and Maxim Rakhuba and Alexander Liniger and Zhiwu Huang and Stamatios Georgoulis and Dengxin Dai and Luc Van Gool},
journal= {arXiv preprint arXiv:2103.04217},
year = {2021}
}
Comments
Accepted at AISTATS 2021