Distributed stochastic gradient descent (SGD) is essential for scaling the machine learning algorithms to a large number of computing nodes. However, the infrastructures variability such as high communication delay or random node slowdown greatly impedes the performance of distributed SGD algorithm, especially in a wireless system or sensor networks. In this paper, we propose an algorithmic approach named Overlap-Local-SGD (and its momentum variant) to overlap the communication and computation so as to speedup the distributed training procedure. The approach can help to mitigate the straggler effects as well. We achieve this by adding an anchor model on each node. After multiple local updates, locally trained models will be pulled back towards the synchronized anchor model rather than communicating with others. Experimental results of training a deep neural network on CIFAR-10 dataset demonstrate the effectiveness of Overlap-Local-SGD. We also provide a convergence guarantee for the proposed algorithm under non-convex objective functions.
The state-of-the-art deep learning algorithms rely on distributed training systems to tackle the increasing sizes of models and training data sets. Minibatch stochastic gradient descent (SGD) algorithm requires workers to halt forward/back propagations, to wait for gradients aggregated from all workers, and to receive weight updates before the next batch of tasks. This synchronous execution model exposes the overheads of gradient/weight communication among a large number of workers in a distributed training system. We propose a new SGD algorithm, DaSGD (Local SGD with Delayed Averaging), which…
Stochastic Gradient Descent (SGD) is the most popular algorithm for training deep neural networks (DNNs). As larger networks and datasets cause longer training times, training on distributed systems is common and distributed SGD variants, mainly asynchronous and synchronous SGD, are widely used. Asynchronous SGD is communication efficient but suffers from accuracy degradation due to delayed parameter updating. Synchronous SGD becomes communication intensive when the number of nodes increases regardless of its advantage. To address these issues, we introduce Layered SGD (LSGD), a new decentrali…
Data parallelism has become the de facto standard for training Deep Neural Network on multiple processing units. In this work we propose DC-S3GD, a decentralized (without Parameter Server) stale-synchronous version of the Delay-Compensated Asynchronous Stochastic Gradient Descent (DC-ASGD) algorithm. In our approach, we allow for the overlap of computation and communication, and compensate the inherent error with a first-order correction of the gradients. We prove the effectiveness of our approach by training Convolutional Neural Network with large batches and achieving state-of-the-art result…
King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）
We revisit the classical problem of finding an approximately stationary point of the average of $n$ smooth and possibly nonconvex functions. The optimal complexity of stochastic first-order methods in terms of the number of gradient evaluations of individual functions is $\mathcal{O}\left(n + n^{1/2}\varepsilon^{-1}\right)$, attained by the optimal SGD methods $\small\sf\color{green}{SPIDER}$(arXiv:1807.01695) and $\small\sf\color{green}{PAGE}$(arXiv:2008.10898), for example, where $\varepsilon$ is the error tolerance. However, i) the big-$\mathcal{O}$ notation hides crucial dependencies on th…
To reduce the long training time of large deep neural network (DNN) models, distributed synchronous stochastic gradient descent (S-SGD) is commonly used on a cluster of workers. However, the speedup brought by multiple workers is limited by the communication overhead. Two approaches, namely pipelining and gradient sparsification, have been separately proposed to alleviate the impact of communication overheads. Yet, the gradient sparsification methods can only initiate the communication after the backpropagation, and hence miss the pipelining opportunity. In this paper, we propose a new distrib…
Modern machine learning frameworks can train neural networks using multiple nodes in parallel, each computing parameter updates with stochastic gradient descent (SGD) and sharing them asynchronously through a central parameter server. Due to communication overhead and bottlenecks, the total throughput of SGD updates in a cluster scales sublinearly, saturating as the number of nodes increases. In this paper, we present a solution to predicting training throughput from profiling traces collected from a single-node configuration. Our approach is able to model the interaction of multiple nodes and…
Neural network learns low-dimensional polynomials with SGD near the information-theoretic limit
Princeton University（普林斯顿大学）；University of California, Berkeley（加利福尼亚大学伯克利分校）；RIKEN AIP（理化学研究所人工智能项目）；University of Tokyo（东京大学）；New York University（纽约大学）；Flatiron Institute（弗拉蒂隆研究所）
Tsinghua University（清华大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
In distributed optimization, a popular technique to reduce communication is quantization. In this paper, we provide a general analysis framework for inexact gradient descent that is applicable to quantization schemes. We also propose a quantization scheme Double Encoding and Error Diminishing (DEED). DEED can achieve small communication complexity in three settings: frequent-communication large-memory, frequent-communication small-memory, and infrequent-communication (e.g. federated learning). More specifically, in the frequent-communication large-memory setting, DEED can be easily combined wi…
Huazhong University of Science and Technology（华中科技大学）；Université de Picardie Jules Verne（皮卡第儒勒-凡尔纳大学）
Stochastic gradient descent (SGD) is a powerful method for large-scale optimization problems in the area of machine learning, especially for a finite-sum formulation with numerous variables. In recent years, mini-batch SGD gains great success and has become a standard technique for training deep neural networks fed with big amount of data. Inspired by its success in deep learning, we apply the idea of SGD with batch selection of samples to a classic optimization problem in decision version. Given $n$ unit circles, the equal circle packing problem (ECPP) asks whether there exist a feasible pack…
Bocconi University（博科尼大学）；Istituto Nazionale di Fisica Nucleare（意大利国家核物理研究所）；Politecnico di Torino（都灵理工大学）；International Centre for Theoretical Physics（国际理论物理中心）
Learning in Deep Neural Networks (DNN) takes place by minimizing a non-convex high-dimensional loss function, typically by a stochastic gradient descent (SGD) strategy. The learning process is observed to be able to find good minimizers without getting stuck in local critical points, and that such minimizers are often satisfactory at avoiding overfitting. How these two features can be kept under control in nonlinear devices composed of millions of tunable connections is a profound and far reaching open question. In this paper we study basic non-convex one- and two-layer neural network models w…
Numerical influence of ReLU'(0) on backpropagation
IRT Saint Exupéry（圣埃克苏佩里研究所）；ISAE-SUPAERO（法国国立高等航空航天学院）；ANITI（图卢兹人工智能跨学科研究所）；Toulouse School of Economics（图卢兹经济学院）；Université Toulouse 1 Capitole（图卢兹第一大学）；Institut de Mathématiques de Toulouse（图卢兹数学研究所）；CNRS（法国国家科学研究中心）；IRIT, Université Paul Sabatier（图卢兹计算机科学研究所，保罗·萨巴捷大学）
INRIA（法国国家信息与自动化研究所）；École Normale Supérieure, Paris（巴黎高等师范学院）；Université de Montréal（蒙特利尔大学）
As datasets continue to increase in size and multi-core computer architectures are developed, asynchronous parallel optimization algorithms become more and more essential to the field of Machine Learning. Unfortunately, conducting the theoretical analysis asynchronous methods is difficult, notably due to the introduction of delay and inconsistency in inherently sequential algorithms. Handling these issues often requires resorting to simplifying but unrealistic assumptions. Through a novel perspective, we revisit and clarify a subtle but important technical issue present in a large fraction of …
Hong Kong University of Science and Technology（香港科技大学）；Microsoft Research（微软研究院）；University of Chicago（芝加哥大学）
Communication overhead poses an important obstacle to distributed DNN training and draws increasing attention in recent years. Despite continuous efforts, prior solutions such as gradient compression/reduction, compute/communication overlapping and layer-wise flow scheduling, etc., are still coarse-grained and insufficient for an efficient distributed training especially when the network is under pressure. We present DLCP, a novel solution exploiting the domain-specific properties of deep learning to optimize communication overhead of DNN training in a fine-grained manner. At its heart, DLCP c…
