How to train your neural ODE: the world of Jacobian and kinetic regularization

Autor:	Finlay, Chris, Jacobsen, Jörn-Henrik, Nurbekyan, Levon, Oberman, Adam M
Rok vydání:	2020
Předmět:	Statistics - Machine Learning Computer Science - Machine Learning
Druh dokumentu:	Working Paper
Popis:	Training neural ODEs on large datasets has not been tractable due to the necessity of allowing the adaptive numerical ODE solver to refine its step size to very small values. In practice this leads to dynamics equivalent to many hundreds or even thousands of layers. In this paper, we overcome this apparent difficulty by introducing a theoretically-grounded combination of both optimal transport and stability regularizations which encourage neural ODEs to prefer simpler dynamics out of all the dynamics that solve a problem well. Simpler dynamics lead to faster convergence and to fewer discretizations of the solver, considerably decreasing wall-clock time without loss in performance. Our approach allows us to train neural ODE-based generative models to the same performance as the unregularized dynamics, with significant reductions in training time. This brings neural ODEs closer to practical relevance in large-scale applications. Comment: Accepted to ICML 2020
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2002.02798 Zobrazit plný text záznamu View this record from Arxiv