Minibatch SGD outperforms Local SGD in heterogeneous distributed learning.
problem Optimizing a combined convex objective with stochastic gradient estimates from different machines.
method Analysis of Minibatch SGD and Local SGD in a heterogeneous distributed setting.
result Minibatch SGD dominates Local SGD in the heterogeneous distributed setting.
Study reveals convergence properties of SGD with random learning rate.
problem Analyzing convergence of SGD with random learning rate in non-convex optimization.
method Introduced Poisson SGD with random learning rate and used stationary distribution analysis.
result Poisson SGD converges to a stationary distribution and finds global minima in non-convex optimization.
Local SGD outperforms minibatch SGD for quadratic objectives.
problem Theoretical foundations of local SGD are lacking.
method Proved local SGD strictly dominates minibatch SGD for quadratic objectives and accelerated local SGD is minimax optimal.
result Local SGD does not dominate minibatch SGD in general convex objectives.
Paper shows local SGD outperforms mini-batch SGD under certain conditions.
problem Proving local SGD's superiority in distributed learning with heterogeneous data.
method New lower and upper bounds for local SGD under first-order heterogeneity assumptions.
result Local SGD is min-max optimal under certain conditions, resolving understanding of distributed optimization.
Overlap-Local-SGD improves distributed SGD by overlapping communication and computation.
problem High communication delay and node slowdown in distributed SGD.
method Adding an anchor model to synchronize local updates and pull them towards the anchor model.
result Overlap-Local-SGD speeds up distributed training and mitigates straggler effects.
New algorithm improves distributed SGD with random sparsification for better convergence and generalization.
problem Communication bottleneck in distributed deep learning.
method Proposes detached error feedback (DEF) algorithm to improve convergence and generalization of communication-efficient distributed SGD.
result Shows better convergence and generalization bounds than existing methods.
Due to its efficiency and ease to implement, stochastic gradient descent (SGD) has been widely used in machine learning. In particular, SGD is one of the most popular optimization methods for distributed learning. Recently, quantized SGD (QSGD), which adopts quantization to reduce the communication cost in SGD-based di…
New bounds for M-SGD show its error distribution is nearly Gaussian.
problem Understanding the error distribution of M-SGD.
method Proved non-asymptotic bounds for M-SGD in Wasserstein distance.
result Error distribution of M-SGD is approximately Gaussian.
UD-SGD analysis shows efficient sampling by a few agents can outperform others.
problem Analyzing convergence speed and sampling strategies in UD-SGD.
method Asymptotic analysis of UD-SGD with various communication patterns and sampling strategies.
result Efficient sampling by a few agents can lead to better overall convergence.
HybridSGD improves SGD performance by balancing computation and communication.
problem Limited scalability and performance of SGD due to communication costs.
method 2D parallel SGD method (HybridSGD) that trades off between 1D s-step SGD and 1D Federated SGD (FedAvg). result HybridSGD achieves better convergence than FedAvg at similar processor scales and up to 121x speedup over FedAvg.
Communication bottleneck has been identified as a significant issue in distributed optimization of large-scale learning models. Recently, several approaches to mitigate this problem have been proposed, including different forms of gradient compression or computing local models and mixing them iteratively. In this paper…
CSER improves SGD efficiency by resetting errors and partial synchronization.
problem Limited scalability of Distributed Stochastic Gradient Descent (SGD) due to communication bottlenecks.
method Introduces 'error reset' technique and partial synchronization for gradients and models.
result Proves convergence for smooth non-convex problems and accelerates distributed training significantly.
Adaptive distributed SGD reduces delay in slow workers.
problem Minimizing delay in distributed SGD with stragglers.
method Adaptive policy for varying k to optimize error-runtime trade-off. result Numerical simulations confirm the effectiveness of the adaptive approach.
STL-SGD accelerates Local SGD by gradually increasing communication periods.
problem Reducing communication complexity in distributed SGD.
method Gradually increasing communication periods while decreasing learning rate.
result Significant reduction in communication complexity with improved convergence rates.
Stochastic gradient descent (SGD) is a popular stochastic optimization method in machine learning. Traditional parallel SGD algorithms, e.g., SimuParallel SGD, often require all nodes to have the same performance or to consume equal quantities of data. However, these requirements are difficult to satisfy when the paral…
SGD-trained neural networks generalize well even with adversarial label noise.
problem Generalization of neural networks trained on adversarial label noise.
method Training a one-hidden-layer neural network with SGD on arbitrary width networks.
result SGD-trained networks achieve classification accuracy competitive with the best halfspace over adversarial label noise.
With the increase in the amount of data and the expansion of model scale, distributed parallel training becomes an important and successful technique to address the optimization challenges. Nevertheless, although distributed stochastic gradient descent (SGD) algorithms can achieve a linear iteration speedup, they are l…
Gradient clipping helps private SGD converge despite potential bias.
problem Gradient clipping in private SGD can bias convergence.
method Theoretical analysis and empirical evaluation of gradient clipping effects.
result Gradient clipping can prevent convergence to stationary points and introduces bias.
Non-convex SGD learns halfspaces with adversarial label noise efficiently.
problem Agnostically learning halfspaces in adversarial label noise settings.
method Non-convex SGD optimization for halfspace learning.
result Non-convex SGD achieves misclassification error close to optimal with adversarial noise.
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 spar…
MindFlayer SGD improves parallel SGD for heterogeneous, random compute times.
problem Minimizing nonconvex functions with heterogeneous, random compute times.
method MindFlayer SGD, designed for stochastic and heterogeneous delays.
result MindFlayer SGD outperforms existing methods in environments with heavy-tailed noise.
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 com…
Machine learning has made tremendous progress in recent years, with models matching or even surpassing humans on a series of specialized tasks. One key element behind the progress of machine learning in recent years has been the ability to train machine learning models in large-scale distributed shared-memory and messa…
Local SGD proves efficient in overparameterized linear regression.
problem Efficiently learning overparameterized linear models in distributed settings.
method Distributed SGD (DSGD) with overparameterized models.
result Excess risk of SGD is smaller than ridge regression in the same sample complexity.
Distributed stochastic gradient descent (SGD) algorithms are widely deployed in training large-scale deep learning models, while the communication overhead among workers becomes the new system bottleneck. Recently proposed gradient sparsification techniques, especially Top-k sparsification with error compensation (To…
A2SGD reduces distributed SGD communication to O(1) per worker.
problem Heavy communication costs in distributed SGD for large models.
method Two-level gradient averaging to consolidate gradients to two local averages.
result Achieves O(1) communication complexity per worker, significantly reducing traffic and training time.
Stochastic Gradient Descent (SGD) is an important algorithm in machine learning. With constant learning rates, it is a stochastic process that, after an initial phase of convergence, generates samples from a stationary distribution. We show that SGD with constant rates can be effectively used as an approximate posterio…
Ringmaster ASGD improves Asynchronous SGD's efficiency under varying worker times.
problem Suboptimal performance of Asynchronous SGD under heterogeneous worker computation times.
method Ringmaster ASGD, a novel Asynchronous SGD method with optimal time complexity.
result Ringmaster ASGD achieves optimal time complexity under arbitrary worker heterogeneity.
Noise balance theory explains SGD's behavior in neural networks.
problem Understanding SGD's navigation in neural network loss landscapes.
method Analyzes minibatch noise and loss function symmetries.
result Derives the stationary distribution of SGD for deep networks.
Unified analysis of asynchronous-SGD algorithms for distributed learning.
problem Analyzing asynchronous-SGD in heterogeneous settings with varying speeds and data distributions.
method Unified convergence theory for non-convex smooth functions, including pure asynchronous SGD and its modifications.
result Unified convergence rates for various asynchronous algorithms, including novel methods.
In this work, we consider the resilience of distributed algorithms based on stochastic gradient descent (SGD) in distributed learning with potentially Byzantine attackers, who could send arbitrary information to the parameter server to disrupt the training process. Toward this end, we propose a new Lipschitz-inspired c…
Local SGD outperforms conventional methods in LLM training.
problem Training large language models on distributed devices.
method Local SGD algorithm applied to distributed training.
result Local SGD achieves competitive results in LLM training.
A distributed SGD method for heterogeneous networks with hubs and workers.
problem Learning in heterogeneous multi-level networks with worker heterogeneity and varying communication.
method Multi-Level Local SGD: distributed SGD with hub-and-spoke paradigm and hub averaging.
result The method converges with error dependent on worker heterogeneity, hub network topology, and iterations.
To accelerate the training of machine learning models, distributed stochastic gradient descent (SGD) and its variants have been widely adopted, which apply multiple workers in parallel to speed up training. Among them, Local SGD has gained much attention due to its lower communication cost. Nevertheless, when the data …
Improves distributed SGD convergence speed with reduced computation load.
problem Mitigating stragglers in distributed SGD to speed up convergence.
method Modeling communication and computation times, adapting number of workers and computation load dynamically.
result Significantly reduces computation load while improving convergence speed.
Communication-efficient SGD algorithms, which allow nodes to perform local updates and periodically synchronize local models, are highly effective in improving the speed and scalability of distributed SGD. However, a rigorous convergence analysis and comparative study of different communication-reduction strategies rem…
Large scale machine learning is increasingly relying on distributed optimization, whereby several machines contribute to the training process of a statistical model. In this work we study the performance of asynchronous, distributed settings, when applying sparsification, a technique used to reduce communication overhe…
SGD converges to an invariant distribution with sub-Gaussian or sub-exponential properties.
problem Optimizing smooth and strongly convex objectives using SGD.
method Analysis through Markov chains, focusing on convergence and concentration properties.
result SGD iterates and their invariant limit distribution inherit sub-Gaussian or sub-exponential concentration properties.
Large-scale distributed training of neural networks is often limited by network bandwidth, wherein the communication time overwhelms the local computation time. Motivated by the success of sketching methods in sub-linear/streaming algorithms, we introduce Sketched SGD, an algorithm for carrying out distributed SGD by c…
Asynchronous SGD can speed up training with a trade-off of gradient staleness.
problem Asynchronous SGD suffers from gradient staleness, affecting convergence error.
method Theoretical analysis of the error-runtime trade-off considering random straggling delays.
result A method of gradually varying synchronicity in distributed SGD is proposed and demonstrated.
Paper improves compressed SGD to reach second-order stationary points.
problem Efficiently reaching second-order stationary points in distributed machine learning.
method Gradient compression and RandomK compressor to improve convergence to second-order stationary points.
result Compressed SGD reaches second-order stationary points with improved communication efficiency.
SGD improves DR by solving two-stage sampling problems.
problem Improving the learning properties of SGD for distribution regression.
method Applying SGD to two-stage sampling problems in distribution regression.
result Theoretical guarantees for SGD's performance in DR, with optimal bounds.
A(DP)2SGD improves federated learning privacy and efficiency.
problem Privacy and efficiency in federated learning with asynchronous decentralized parallel SGD.
method Differentially private asynchronous decentralized parallel SGD (A(DP)2SGD) using R{é}nyi differential privacy. result Achieves optimal convergence rate and comparable model accuracy to SSGD but faster.
The paper investigates heavy-tailed behavior in offline SGD, showing it approximates power-law tails.
problem Understanding heavy-tailed behavior in offline (multi-pass) SGD with finite data.
method Proves nonasymptotic Wasserstein convergence bounds for offline SGD to online SGD.
result Offline SGD exhibits approximate power-law tails as the number of data points increases.
Distributing Neural Network training is of particular interest for several reasons including scaling using computing clusters, training at data sources such as IOT devices and edge servers, utilizing underutilized resources across heterogeneous environments, and so on. Most contemporary approaches primarily address sca…
New convergence bounds for shuffling-based SGD methods in distributed learning.
problem Analyzing the performance of shuffling-based variants of SGD in distributed learning.
method Study of minibatch and local Random Reshuffling methods, proving convergence bounds and lower bounds.
result Shuffling-based variants converge faster than with-replacement sampling methods, and the bounds are tight.
Stochastic Gradient Descent (SGD) has become one of the most popular optimization methods for training machine learning models on massive datasets. However, SGD suffers from two main drawbacks: (i) The noisy gradient updates have high variance, which slows down convergence as the iterates approach the optimum, and (ii)…
New methods improve distributed optimization on non-iid data.
problem Communication bottleneck in distributed machine learning models.
method Two types of distributed gradient compression methods (D-QSGD and D-EF-SGD) analyzed for non-iid data.
result D-EF-SGD performs better than D-QSGD on non-iid data but can still slow down with high data skewness.