Shampoo: Preconditioned Stochastic Tensor Optimization
Vineet Gupta, Tomer Koren, Yoram Singer
Abstract
Preconditioned gradient methods are among the most general and powerful tools in optimization. However, preconditioning requires storing and manipulating prohibitively large matrices. We describe and analyze a new structure-aware preconditioning algorithm, called Shampoo, for stochastic optimization over tensor spaces. Shampoo maintains a set of preconditioning matrices, each of which operates on a single dimension, contracting over the remaining dimensions. We establish convergence guarantees in the stochastic convex setting, the proof of which builds upon matrix trace inequalities. Our experiments with state-of-the-art deep learning models show that Shampoo is capable of converging considerably faster than commonly used optimizers. Surprisingly, although it involves a more complex update rule, Shampoo’s runtime per step is comparable in practice to that of simple gradient methods such as SGD, AdaGrad, and Adam.
BibTeX
@InProceedings{pmlr-v80-gupta18a,
title = {Shampoo: Preconditioned Stochastic Tensor Optimization},
author = {Gupta, Vineet and Koren, Tomer and Singer, Yoram},
booktitle = {Proceedings of the 35th International Conference on Machine Learning},
pages = {1842--1850},
year = {2018},
editor = {Dy, Jennifer and Krause, Andreas},
volume = {80},
series = {Proceedings of Machine Learning Research},
month = {10--15 Jul},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v80/gupta18a/gupta18a.pdf},
url = {https://proceedings.mlr.press/v80/gupta18a.html},
abstract = {Preconditioned gradient methods are among the most general and powerful tools in optimization. However, preconditioning requires storing and manipulating prohibitively large matrices. We describe and analyze a new structure-aware preconditioning algorithm, called Shampoo, for stochastic optimization over tensor spaces. Shampoo maintains a set of preconditioning matrices, each of which operates on a single dimension, contracting over the remaining dimensions. We establish convergence guarantees in the stochastic convex setting, the proof of which builds upon matrix trace inequalities. Our experiments with state-of-the-art deep learning models show that Shampoo is capable of converging considerably faster than commonly used optimizers. Surprisingly, although it involves a more complex update rule, Shampoo’s runtime per step is comparable in practice to that of simple gradient methods such as SGD, AdaGrad, and Adam.}
}