We provide an improved analysis of normalized SGD showing that adding momentum provably removes the need for large batch sizes on non-convex objectives. Then, we consider the case of objectives with bounded second derivative and show that in this case a small tweak to the momentum formula allows normalized SGD with momentum to find an $ε$-critical point in $O(1/ε^{3.5})$ iterations, matching the best-known rates without accruing any logarithmic factors or dependence on dimension. We also provide an adaptive method that automatically improves convergence rates when the variance in the gradients is small. Finally, we show that our method is effective when employed on popular large scale tasks such as ResNet-50 and BERT pretraining, matching the performance of the disparate methods used to get state-of-the-art results on both tasks.
Toyota Technological Institute at Chicago（芝加哥丰田理工学院）；University of Toronto（多伦多大学）
We revisit the choice of SGD for training deep neural networks by reconsidering the appropriate geometry in which to optimize the weights. We argue for a geometry invariant to rescaling of weights that does not affect the output of the network, and suggest Path-SGD, which is an approximate steepest descent method with respect to a path-wise regularizer related to max-norm regularization. Path-SGD is easy and efficient to implement and leads to empirical gains over SGD and AdaGrad.
University of Calgary（卡尔加里大学）；Wuhan University（武汉大学）；The University of Hong Kong（香港大学）
Adaptive optimization algorithms, such as Adam and RMSprop, have shown better optimization performance than stochastic gradient descent (SGD) in some scenarios. However, recent studies show that they often lead to worse generalization performance than SGD, especially for training deep neural networks (DNNs). In this work, we identify the reasons that Adam generalizes worse than SGD, and develop a variant of Adam to eliminate the generalization gap. The proposed method, normalized direction-preserving Adam (ND-Adam), enables more precise control of the direction and step size for updating weigh…
While momentum-based methods, in conjunction with stochastic gradient descent (SGD), are widely used when training machine learning models, there is little theoretical understanding on the generalization error of such methods. In this work, we first show that there exists a convex loss function for which algorithmic stability fails to establish generalization guarantees when SGD with standard heavy-ball momentum (SGDM) is run for multiple epochs. Then, for smooth Lipschitz loss functions, we analyze a modified momentum-based update rule, i.e., SGD with early momentum (SGDEM), and show that it …
Momentum methods are now used pervasively within the machine learning community for training non-convex models such as deep neural networks. Empirically, they out perform traditional stochastic gradient descent (SGD) approaches. In this work we develop a Lyapunov analysis of SGD with momentum (SGD+M), by utilizing a equivalent rewriting of the method known as the stochastic primal averaging (SPA) form. This analysis is much tighter than previous theory in the non-convex case, and due to this we are able to give precise insights into when SGD+M may out-perform SGD, and what hyper-parameter sche…
Boston University（波士顿大学）；George Mason University（乔治梅森大学）
SGD with Momentum (SGDM) is a widely used family of algorithms for large-scale optimization of machine learning problems. Yet, when optimizing generic convex functions, no advantage is known for any SGDM algorithm over plain SGD. Moreover, even the most recent results require changes to the SGDM algorithms, like averaging of the iterates and a projection onto a bounded domain, which are rarely used in practice. In this paper, we focus on the convergence rate of the last iterate of SGDM. For the first time, we prove that for any constant momentum factor, there exists a Lipschitz and convex func…
Mila - Quebec AI Institute & Université de Montréal（Mila - 魁北克人工智能研究所 & 蒙特利尔大学）；Google DeepMind（谷歌DeepMind）；Massachusetts Institute of Technology (MIT)（麻省理工学院）；McGill University（麦吉尔大学）
University of California, Los Angeles（加利福尼亚大学洛杉矶分校）；US Army Research Lab（美国陆军研究实验室）
In this paper, we propose and analyze SQuARM-SGD, a communication-efficient algorithm for decentralized training of large-scale machine learning models over a network. In SQuARM-SGD, each node performs a fixed number of local SGD steps using Nesterov's momentum and then sends sparsified and quantized updates to its neighbors regulated by a locally computable triggering criterion. We provide convergence guarantees of our algorithm for general (non-convex) and convex smooth objectives, which, to the best of our knowledge, is the first theoretical analysis for compressed decentralized SGD with mo…
Recent developments on large-scale distributed machine learning applications, e.g., deep neural networks, benefit enormously from the advances in distributed non-convex optimization techniques, e.g., distributed Stochastic Gradient Descent (SGD). A series of recent works study the linear speedup property of distributed SGD variants with reduced communication. The linear speedup property enable us to scale out the computing capability by adding more computing nodes into our system. The reduced communication complexity is desirable since communication overhead is often the performance bottleneck…
In this paper, we propose a novel accelerated stochastic gradient method with momentum, which momentum is the weighted average of previous gradients. The weights decays inverse proportionally with the iteration times. Stochastic gradient descent with momentum (Sgdm) use weights that decays exponentially with the iteration times to generate an momentum term. Using exponentially decaying weights, variants of Sgdm with well designed and complicated formats have been proposed to achieve better performance. The momentum update rules of our method is as simple as that of Sgdm. We provide theoretical…
Toyota Technological Institute at Chicago（芝加哥丰田技术研究院）；University of Toronto（多伦多大学）；Carnegie Mellon University（卡内基梅隆大学）
We investigate the parameter-space geometry of recurrent neural networks (RNNs), and develop an adaptation of path-SGD optimization method, attuned to this geometry, that can learn plain RNNs with ReLU activations. On several datasets that require capturing long-term dependency structure, we show that path-SGD can significantly improve trainability of ReLU RNNs compared to RNNs trained with SGD, even with various recently suggested initialization schemes.
School of Industrial and Systems Engineering at Georgia Tech（佐治亚理工学院工业与系统工程学院）
Momentum Stochastic Gradient Descent (MSGD) algorithm has been widely applied to many nonconvex optimization problems in machine learning, e.g., training deep neural networks, variational Bayesian inference, and etc. Despite its empirical success, there is still a lack of theoretical understanding of convergence properties of MSGD. To fill this gap, we propose to analyze the algorithmic behavior of MSGD by diffusion approximations for nonconvex optimization problems with strict saddle points and isolated local optima. Our study shows that the momentum helps escape from saddle points, but hurts…
State Key Laboratory of Scientific and Engineering Computing（科学与工程计算国家重点实验室）；Academy of Mathematics and Systems Science, Chinese Academy of Sciences（中国科学院数学与系统科学研究院）
Following the same routine as [SSJ20], we continue to present the theoretical analysis for stochastic gradient descent with momentum (SGD with momentum) in this paper. Differently, for SGD with momentum, we demonstrate it is the two hyperparameters together, the learning rate and the momentum coefficient, that play the significant role for the linear rate of convergence in non-convex optimization. Our analysis is based on the use of a hyperparameters-dependent stochastic differential equation (hp-dependent SDE) that serves as a continuous surrogate for SGD with momentum. Similarly, we establis…
Stochastic Gradient Descent (SGD) and its variants are the most used algorithms in machine learning applications. In particular, SGD with adaptive learning rates and momentum is the industry standard to train deep networks. Despite the enormous success of these methods, our theoretical understanding of these variants in the nonconvex setting is not complete, with most of the results only proving convergence in expectation and with strong assumptions on the stochastic gradients. In this paper, we present a high probability analysis for adaptive and momentum algorithms, under weak assumptions on…
Acceleration of stochastic gradient descent with momentum by averaging: finite-sample rates and asymptotic normality
School of Mathematical Sciences, Shanghai Jiao Tong University（上海交通大学数学科学学院）；Daniels School of Business, Purdue University（普渡大学丹尼尔斯商学院）；Stern School of Business, New York University（纽约大学斯特恩商学院）
Most convergence guarantees for stochastic gradient descent with momentum (SGDm) rely on iid sampling. Yet, SGDm is often used outside this regime, in settings with temporally correlated input samples such as continual learning and reinforcement learning. Existing work has shown that SGDm with a decaying step-size can converge under Markovian temporal correlation. In this work, we show that SGDm under covariate shift with a fixed step-size can be unstable and diverge. In particular, we show SGDm under covariate shift is a parametric oscillator, and so can suffer from a phenomenon known as reso…
University of Science and Technology of China（中国科学技术大学）；Microsoft Research Asia（微软亚洲研究院）
By ensuring differential privacy in the learning algorithms, one can rigorously mitigate the risk of large models memorizing sensitive training data. In this paper, we study two algorithms for this purpose, i.e., DP-SGD and DP-NSGD, which first clip or normalize \textit{per-sample} gradients to bound the sensitivity and then add noise to obfuscate the exact information. We analyze the convergence behavior of these two algorithms in the non-convex optimization setting with two common assumptions and achieve a rate $\mathcal{O}\left(\sqrt[4]{\frac{d\log(1/δ)}{N^2ε^2}}\right)$ of the gradient nor…
We develop a new algorithm for non-convex stochastic optimization that finds an $ε$-critical point in the optimal $O(ε^{-3})$ stochastic gradient and Hessian-vector product computations. Our algorithm uses Hessian-vector products to "correct" a bias term in the momentum of SGD with momentum. This leads to better gradient estimates in a manner analogous to variance reduction methods. In contrast to prior work, we do not require excessively large batch sizes, and are able to provide an adaptive algorithm whose convergence rate automatically improves with decreasing variance in the gradient estim…
We analyze the dynamics of large batch stochastic gradient descent with momentum (SGD+M) on the least squares problem when both the number of samples and dimensions are large. In this setting, we show that the dynamics of SGD+M converge to a deterministic discrete Volterra equation as dimension increases, which we analyze. We identify a stability measurement, the implicit conditioning ratio (ICR), which regulates the ability of SGD+M to accelerate the algorithm. When the batch size exceeds this ICR, SGD+M converges linearly at a rate of $\mathcal{O}(1/\sqrtκ)$, matching optimal full-batch mome…
Nesterov SGD is widely used for training modern neural networks and other machine learning models. Yet, its advantages over SGD have not been theoretically clarified. Indeed, as we show in our paper, both theoretically and empirically, Nesterov SGD with any parameter selection does not in general provide acceleration over ordinary SGD. Furthermore, Nesterov SGD may diverge for step sizes that ensure convergence of ordinary SGD. This is in contrast to the classical results in the deterministic scenario, where the same step size ensures accelerated convergence of the Nesterov's method over optim…
Amazon Web Services（亚马逊云科技）；Hong Kong University of Science and Technology（香港科技大学）
Communication overhead is a major bottleneck hampering the scalability of distributed machine learning systems. Recently, there has been a surge of interest in using gradient compression to improve the communication efficiency of distributed neural network training. Using 1-bit quantization, signSGD with majority vote achieves a 32x reduction on communication cost. However, its convergence is based on unrealistic assumptions and can diverge in practice. In this paper, we propose a general distributed compressed SGD with Nesterov's momentum. We consider two-way compression, which compresses the…
High-dimensional limit theorems for SGD: Momentum and Adaptive Step-sizes
Department of Statistics and Actuarial Science, University of Waterloo, Canada（统计与精算科学系，滑铁卢大学，加拿大）；Cheriton School of Computer Science, University of Waterloo, Canada（切尔顿计算机科学学院，滑铁卢大学，加拿大）
Variance reduction has emerged in recent years as a strong competitor to stochastic gradient descent in non-convex problems, providing the first algorithms to improve upon the converge rate of stochastic gradient descent for finding first-order critical points. However, variance reduction techniques typically require carefully tuned learning rates and willingness to use excessively large "mega-batches" in order to achieve their improved results. We present a new algorithm, STORM, that does not require any batches and makes use of adaptive learning rates, enabling simpler implementation and les…
On the Generalization of Stochastic Gradient Descent with Momentum
University of Oslo（奥斯陆大学）；Visual Intelligence Centre（视觉智能中心）；Integreat, Norwegian Centre for Knowledge-driven Machine Learning（Integreat，挪威知识驱动机器学习中心）；EPFL（洛桑联邦理工学院）；Laboratory for Information and Inference Systems (LIONS), EPFL（信息与推理系统实验室（LIONS），洛桑联邦理工学院）；University of Toronto（多伦多大学）
Alibaba Group（阿里巴巴集团）；University of Pennsylvania（宾夕法尼亚大学）
The scale of deep learning nowadays calls for efficient distributed training algorithms. Decentralized momentum SGD (DmSGD), in which each node averages only with its neighbors, is more communication efficient than vanilla Parallel momentum SGD that incurs global average across all computing nodes. On the other hand, the large-batch training has been demonstrated critical to achieve runtime speedup. This motivates us to investigate how DmSGD performs in the large-batch scenario. In this work, we find the momentum term can amplify the inconsistency bias in DmSGD. Such bias becomes more evident …
When and Why Momentum Accelerates SGD:An Empirical Study
University of Science and Technology of China（中国科学技术大学）；Microsoft Research（微软研究院）；Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）
A theoretical, and potentially also practical, problem with stochastic gradient descent is that trajectories may escape to infinity. In this note, we investigate uniform boundedness properties of iterates and function values along the trajectories of the stochastic gradient descent algorithm and its important momentum variant. Under smoothness and $R$-dissipativity of the loss function, we show that broad families of step-sizes, including the widely used step-decay and cosine with (or without) restart step-sizes, result in uniformly bounded iterates and function values. Several important appli…
National University of Defense Technology（国防科技大学）；Tsinghua University（清华大学）；University of Utah（犹他大学）
Heavy ball momentum is crucial in accelerating (stochastic) gradient-based optimization algorithms for machine learning. Existing heavy ball momentum is usually weighted by a uniform hyperparameter, which relies on excessive tuning. Moreover, the calibrated fixed hyperparameter may not lead to optimal performance. In this paper, to eliminate the effort for tuning the momentum-related hyperparameter, we propose a new adaptive momentum inspired by the optimal choice of the heavy ball momentum for quadratic optimization. Our proposed adaptive heavy ball momentum can improve stochastic gradient de…
University of California, Berkeley（加州大学伯克利分校）；Queensland University of Technology（昆士兰科技大学）；Cornell University（康奈尔大学）；University of Illinois, Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
This paper presents a margin-based multiclass generalization bound for neural networks that scales with their margin-normalized "spectral complexity": their Lipschitz constant, meaning the product of the spectral norms of the weight matrices, times a certain correction factor. This bound is empirically investigated for a standard AlexNet network trained with SGD on the mnist and cifar10 datasets, with both original and random labels; the bound, the Lipschitz constants, and the excess risks are all in direct correlation, suggesting both that SGD selects predictors whose complexity scales with t…
Carnegie Mellon University（卡内基梅隆大学）；Facebook AI Research（脸书人工智能研究院）
Distributed optimization is essential for training large models on large datasets. Multiple approaches have been proposed to reduce the communication overhead in distributed training, such as synchronizing only after performing multiple local SGD steps, and decentralized methods (e.g., using gossip algorithms) to decouple communications among workers. Although these methods run faster than AllReduce-based methods, which use blocking communication before every update, the resulting models may be less accurate after the same number of updates. Inspired by the BMUF method of Chen & Huo (2016), we…
Error estimates between SGD with momentum and underdamped Langevin diffusion
Laboratoire de Mathematiques Blaise Pascal, CNRS-UMR 6620, Universite Clermont-Auvergne（克莱蒙奥弗涅大学）；University of Macau（澳门大学）；UM Zhuhai Research Institute（澳门大学珠海研究院）；School of Mathematical Sciences, Peking University（北京大学数学科学学院）；Beijing International Center for Mathematical Research (BICMR), Peking University（北京大学北京国际数学研究中心）
The University of Tokyo（东京大学）；RIKEN Center for AIP（理化学研究所人工智能研究中心）；Microsoft Research Asia（微软亚洲研究院）
Adaptive Moment Estimation (Adam), which combines Adaptive Learning Rate and Momentum, would be the most popular stochastic optimizer for accelerating the training of deep neural networks. However, it is empirically known that Adam often generalizes worse than Stochastic Gradient Descent (SGD). The purpose of this paper is to unveil the mystery of this behavior in the diffusion theoretical framework. Specifically, we disentangle the effects of Adaptive Learning Rate and Momentum of the Adam dynamics on saddle-point escaping and flat minima selection. We prove that Adaptive Learning Rate can es…
Courant Institute of Mathematical Sciences, New York University（纽约大学库朗数学科学研究所）
When solving finite-sum minimization problems, two common alternatives to stochastic gradient descent (SGD) with theoretical benefits are random reshuffling (SGD-RR) and shuffle-once (SGD-SO), in which functions are sampled in cycles without replacement. Under a convenient stochastic noise approximation which holds experimentally, we study the stationary variances of the iterates of SGD, SGD-RR and SGD-SO, whose leading terms decrease in this order, and obtain simple approximations. To obtain our results, we study the power spectral density of the stochastic gradient noise sequences. Our analy…
LTCI, Télécom Paris, Institut Polytechnique de Paris（巴黎理工学院巴黎电信学院LTCI实验室）；Department of Statistics, University of Oxford（牛津大学统计系）；Department of Mathematics, Florida State University（佛罗里达州立大学数学系）；Department of Management Science and Information Systems, Rutgers Business School（罗格斯大学商学院管理科学与信息系统系）
Stochastic gradient descent with momentum (SGDm) is one of the most popular optimization algorithms in deep learning. While there is a rich theory of SGDm for convex problems, the theory is considerably less developed in the context of deep learning where the problem is non-convex and the gradient noise might exhibit a heavy-tailed behavior, as empirically observed in recent studies. In this study, we consider a \emph{continuous-time} variant of SGDm, known as the underdamped Langevin dynamics (ULD), and investigate its asymptotic properties under heavy-tailed perturbations. Supported by recen…
Microsoft（微软）；University of Rochester（罗切斯特大学）；ETH Zurich（苏黎世联邦理工学院）
Adam is the important optimization algorithm to guarantee efficiency and accuracy for training many important tasks such as BERT and ImageNet. However, Adam is generally not compatible with information (gradient) compression technology. Therefore, the communication usually becomes the bottleneck for parallelizing Adam. In this paper, we propose a communication efficient {\bf A}DAM {\bf p}reconditioned {\bf M}omentum SGD algorithm-- named APMSqueeze-- through an error compensated method compressing gradients. The proposed algorithm achieves a similar convergence efficiency to Adam in term of ep…
Facebook AI Research（Facebook人工智能研究院）；New York University（纽约大学）
Momentum-based acceleration of stochastic gradient descent (SGD) is widely used in deep learning. We propose the quasi-hyperbolic momentum algorithm (QHM) as an extremely simple alteration of momentum SGD, averaging a plain SGD step with a momentum step. We describe numerous connections to and identities with other algorithms, and we characterize the set of two-state optimization algorithms that QHM can recover. Finally, we propose a QH variant of Adam called QHAdam, and we empirically demonstrate that our algorithms lead to significantly improved training in a variety of settings, including a…
In stochastic gradient descent, especially for neural network training, there are currently dominating first order methods: not modeling local distance to minimum. This information required for optimal step size is provided by second order methods, however, they have many difficulties, starting with full Hessian having square of dimension number of coefficients. This article proposes a minimal step from successful first order momentum method toward second order: online parabola modelling in just a single direction: normalized $\hat{v}$ from momentum method. It is done by estimating linear tren…
DNT: a Deeply Normalized Transformer that can be trained by Momentum SGD
Intellifusion Inc.（Intellifusion公司）；Tsinghua University（清华大学）；Johns Hopkins University（约翰霍普金斯大学）；Beijing University of Posts and Telecommunications（北京邮电大学）；Peking University（北京大学）
本文提出深度归一化 Transformer（DNT），通过在网络关键位置加入归一化来调节 Jacobian 矩阵并集中梯度分布，使普通 mSGDW 也能有效训练，且在 ViT 与 GPT 上表现更优。
A matrix free and a low rank approximation preconditioner are proposed to accelerate the convergence of stochastic gradient descent (SGD) by exploiting curvature information sampled from Hessian-vector products or finite differences of parameters and gradients similar to the BFGS algorithm. Both preconditioners are fitted with an online updating manner minimizing a criterion that is free of line search and robust to stochastic gradient noise, and further constrained to be on certain connected Lie groups to preserve their corresponding symmetry or invariance, e.g., orientation of coordinates by…
