Search

In deep learning, different kinds of deep networks typically need different optimizers, which have to be chosen after multiple trials, making the training process inefficient. To relieve this issue and consistently improve the model training speed across deep networks, we propose the ADAptive Nesterov momentum algorithm, Adan for short. Adan first reformulates the vanilla Nesterov acceleration to develop a new Nesterov momentum estimation (NME) method, which avoids the extra overhead of computing gradient at the extrapolation point. Then Adan adopts NME to estimate the gradient's first- and second-order moments in adaptive gradient algorithms for convergence acceleration. Besides, we prove that Adan finds an ϵ-approximate first-order stationary point within O(ϵ -3.5 ) stochastic gradient complexity on the non-convex stochastic problems (e.g.deep learning problems), matching the best-known lower bound. Extensive experimental results show that Adan consistently surpasses the corresponding SoTA optimizers on vision, language, and RL tasks and sets new SoTAs for many popular networks and frameworks, eg ResNet, ConvNext, ViT, Swin, MAE, DETR, GPT-2, Transformer-XL, and BERT. More surprisingly, Adan can use half of the training cost (epochs) of SoTA optimizers to achieve higher or comparable performance on ViT, GPT-2, MAE, etc, and also shows great tolerance to a large range of minibatch size, e.g.from 1k to 32k. Code is released at https://github.com/sail-sg/Adan, and has been used in multiple popular deep learning frameworks or projects.

Despite the success of stochastic variance-reduced gradient (SVRG) algorithms in solving large-scale problems, their stochastic gradient complexity often scales linearly with data size and is expensive for huge data. Accordingly, we propose a hybrid stochastic-deterministic minibatch proximal gradient~(HSDMPG) algorithm for strongly convex problems with linear prediction structure, e.g.~least squares and logistic/softmax regression. HSDMPG~enjoys improved computational complexity that is data-size-independent for large-scale problems. It iteratively samples an evolving~minibatch of individual losses to estimate the original problem, and efficiently minimizes the sampled smaller-sized subproblems. For strongly convex loss of n components, HSDMPG~attains an ϵ-optimization-error within [Formula: see text] stochastic gradient evaluations, where κ is condition number, ζ = 1 for quadratic loss and ζ = 2 for generic loss. For large-scale problems, our complexity outperforms those of SVRG-type algorithms with/without dependence on data size. Particularly, when ϵ = O(1/√n) which matches the intrinsic excess error of a learning model and is sufficient for generalization, our complexity for quadratic and generic losses is respectively O (n 0.5 log 2 (n)) and O (n 0.5 log 3 (n)), which for the first time achieves optimal generalization in less than a single pass over data. Besides, we extend HSDMPG~to online strongly convex problems and prove its higher efficiency over the prior algorithms. Numerical results demonstrate the computational advantages of~HSDMPG.