We propose ACProp (Asynchronous-centering-Prop), an adaptive optimizer which combines centering of second momentum and asynchronous update (e.g. for $t$-th update, denominator uses information up to step $t-1$, while numerator uses gradient at $t$-th step). ACProp has both strong theoretical properties and empirical performance. With the example by Reddi et al. (2018), we show that asynchronous optimizers (e.g. AdaShift, ACProp) have weaker convergence condition than synchronous optimizers (e.g. Adam, RMSProp, AdaBelief); within asynchronous optimizers, we show that centering of second momentum further weakens the convergence condition. We demonstrate that ACProp has a convergence rate of $O(\frac{1}{\sqrt{T}})$ for the stochastic non-convex case, which matches the oracle rate and outperforms the $O(\frac{logT}{\sqrt{T}})$ rate of RMSProp and Adam. We validate ACProp in extensive empirical studies: ACProp outperforms both SGD and other adaptive optimizers in image classification with CNN, and outperforms well-tuned adaptive optimizers in the training of various GAN models, reinforcement learning and transformers. To sum up, ACProp has good theoretical properties including weak convergence condition and optimal convergence rate, and strong empirical performance including good generalization like SGD and training stability like Adam. We provide the implementation at https://github.com/juntang-zhuang/ACProp-Optimizer.

翻译：我们提出ACPROp(Sycron-center-centering-Prop),这是一个适应性优化机制,它将第二次势头和不同步更新(例如,美元-美元更新,分母使用信息到美元-美元,分子使用梯度为美元-美元-美元-第一级)。ACPROp既具有很强的理论性能,也具有很强的经验性能。Reddi等人(2018年)的例子显示,不同步优化机制(例如,AdaShift、ACProp)比同步优化机制(例如,Adam,RMSProp,Adabelief)的趋同性能更弱;在不同步优化的优化机制内,我们显示,第二动力集中使用信息到美元-美元-美元,而分子使用梯度梯度(fercoc{1unsqrt{Tangangangang) 。我们表明,ACPFACP的不稳性优化非软性优化模式(例如,OFAS-rmalmodal-rmalal-ration Stabilational-rmalation Stabilation Stabilation Stabilation) 和ASlimst Stabilevidalizalizalizalizal ASildaldaldalizaldalizalizalizaldalizaldaldaldaldalizaldaldaldaldaldaldaldal:A:As ASlation ASlations ASlation:Acalticalticalticaldaldaldaldalticaldaldaldald. AS ASmaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldaldald