Asynchronous parallel adaptive stochastic gradient methods

Stochastic gradient methods (SGMs) are the predominant approaches to train deep learning models. The adaptive versions (e.g., Adam and AMSGrad) have been extensively used in practice, partly because they achieve faster convergence than the non-adaptive versions while incurring little overhead. On the other hand, asynchronous (async) parallel computing has exhibited much better speed-up over its synchronous (sync) counterpart. However, async-parallel implementation has only been demonstrated to the non-adaptive SGMs. The difficulty for adaptive SGMs originates from the second moment term that makes the convergence analysis challenging with async updates. In this paper, we propose an async-parallel adaptive SGM based on AMSGrad. We show that the proposed method inherits the convergence guarantee of AMSGrad for both convex and non-convex problems, if the staleness (also called delay) caused by asynchrony is bounded. Our convergence rate results indicate a nearly linear parallelization speed-up if $\tau=o(K^{\frac{1}{4}})$, where $\tau$ is the staleness and $K$ is the number of iterations. The proposed method is tested on both convex and non-convex machine learning problems, and the numerical results demonstrate its clear advantages over the sync counterpart.