Large batch size training of Neural Networks has been shown to incur accuracy loss when trained with the current methods. The exact underlying reasons for this are still not completely understood. Here, we study large batch size training through the lens of the Hessian operator and robust optimization. In particular, w…
On-device research index
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
168,657 papers · 148 categories
Trend · papers per month
48 results for “large-batch operations”
SNGM improves large-batch training accuracy.
problem Improving generalization in large-batch training.
method Stochastic Normalized Gradient Descent with Momentum.
result SNGM achieves better test accuracy than MSGD and other large-batch methods.
Introduces a differentiable approximation to the zero-one loss.
problem Incompatibility of zero-one loss with gradient-based optimization.
method Smooth projection onto hypersimplex through constrained optimization.
result Achieves significant improvements in generalization under large-batch training.
SVRN accelerates Newton methods by reducing variance and improving performance.
problem Improving the efficiency of Newton methods for large-scale optimization problems.
method Stochastic Variance-Reduced Newton (SVRN) algorithm that accelerates Subsampled Newton and Iterative Hessian Sketch algorithms.
result SVRN accelerates Newton methods by reducing the number of passes over the data, achieving a significant improvement in performance.
DReg boosts large-batch SGD's generalization and convergence.
problem Large-batch SGD struggles with generalization in deep learning.
method DReg replicates a layer to encourage parameter diversity.
result DReg improves generalization and convergence with large-batch SGD.
Training deep neural networks using a large batch size has shown promising results and benefits many real-world applications. However, the optimizer converges slowly at early epochs and there is a gap between large-batch deep learning optimization heuristics and theoretical underpinnings. In this paper, we propose a no…
Distributed implementations of mini-batch stochastic gradient descent (SGD) suffer from communication overheads, attributed to the high frequency of gradient updates inherent in small-batch training. Training with large batches can reduce these overheads; however, large batches can affect the convergence properties and…
Large batch sizes reduce gradient variance in DP-SGD, improving privacy.
problem Understanding why large batch sizes work in DP-SGD.
method Decomposed total gradient variance into subsampling and noise-induced variances, proving batch size independence in the limit.
result Large batch sizes reduce effective total gradient variance, improving privacy in DP-SGD.
In stochastic optimization, using large batch sizes during training can leverage parallel resources to produce faster wall-clock training times per training epoch. However, for both training loss and testing error, recent results analyzing large batch Stochastic Gradient Descent (SGD) have found sharp diminishing retur…
Large-batch training approaches have enabled researchers to utilize large-scale distributed processing and greatly accelerate deep-neural net (DNN) training. For example, by scaling the batch size from 256 to 32K, researchers have been able to reduce the training time of ResNet50 on ImageNet from 29 hours to 2.2 minute…
New attack recovers user-level information from large batch images.
problem Recovering private information from user-level gradients in distributed learning.
method Proposes a gradient inversion attack using a denoising diffusion model as a prior.
result Demonstrates recovery of realistic facial images and private attributes.
A Large Batch Optimizer Reality Check: Traditional, Generic Optimizers Suffice Across Batch Sizescs.LG
Standard optimizers perform as well as LARS and LAMB at large batch sizes.
problem Comparing optimizers for neural network training at large batch sizes.
method Used standard optimizers like Nesterov momentum and Adam to match or exceed LARS and LAMB results.
result Standard optimizers can match or exceed LARS and LAMB at large batch sizes.
Large batch training with DP-SGD reduces model performance due to implicit bias.
problem Large batch training with DP-SGD reduces model performance.
method The study analyzes the phenomenon of implicit bias in Noisy-SGD (DP-SGD without clipping) and its theoretical solutions for linear models.
result The implicit bias in large batch training with DP-SGD is amplified by additional noise, similar to SGD.
AdaScale SGD adapts learning rates for large-batch training efficiently.
problem Adapting learning rates for large-batch training to balance speed-ups and model quality.
method Adaptive learning rate adaptation based on gradient variance.
result AdaScale achieves reliable speed-ups for a wide range of batch sizes without degrading model quality.
LSAM optimizes deep learning training with improved efficiency.
problem Inefficiency in distributed large-batch training with Sharpness-Aware Minimization (SAM).
method Integrates SAM's adversarial steps with an asynchronous distributed sampling strategy.
result Higher final accuracy compared to data-parallel SAM.
Noise in SGD helps deep nets generalize better, even with smaller batch sizes.
problem The generalization benefit of using noise in SGD over large batch sizes.
method Carefully designed experiments and rigorous hyperparameter sweeps on various models.
result Small or moderately large batch sizes outperform very large batches on test sets.
Generative models enhance BO for large batch optimization.
problem Efficiently sampling solutions in high-dimensional, combinatorial design spaces.
method Train generative models to sample solutions proportional to expected utility.
result Generative models can approximate optimal target distributions under certain conditions.
Variance reduction methods such as SVRG and SpiderBoost use a mixture of large and small batch gradients to reduce the variance of stochastic gradients. Compared to SGD, these methods require at least double the number of operations per update to model parameters. To reduce the computational cost of these methods, we i…
Momentum affects optimization differently at small vs large batch sizes near instability.
problem Understanding how momentum impacts optimization near the edge of stability.
method Demonstrated through batch-size dependent behavior of SGD with momentum.
result Momentum operates in two distinct regimes: amplifying stochastic fluctuations at small batch sizes and stabilizing at large batch sizes.
New framework optimizes deep learning training by deferring large batch sizes to late stages.
problem Optimizing batch size scheduling for deep learning training efficiency.
method Introduced the functional scaling law (FSL) framework to analyze and optimize batch size scheduling.
result Large batch sizes can be deferred to late training stages without sacrificing performance.
This paper explores the limits of large batch sizes in deep learning.
problem Understanding the optimal batch size for deep learning models.
method Detailed numerical optimization and experimental analysis of large batch sizes.
result Improvement of 18% top-1 test accuracy with an optimized batch size recipe.
Muon optimizes training efficiency by improving data retention at large batch sizes.
problem Improving training efficiency and data retention at large batch sizes.
method Introducing Muon, a second-order optimizer, and combining it with muP for efficient hyperparameter transfer.
result Muon outperforms AdamW in retaining data efficiency at large batch sizes, enabling more economical training.
The choice of batch-size in a stochastic optimization algorithm plays a substantial role for both optimization and generalization. Increasing the batch-size used typically improves optimization but degrades generalization. To address the problem of improving generalization while maintaining optimal convergence in large…
Training large deep neural networks on massive datasets is computationally very challenging. There has been recent surge in interest in using large batch stochastic optimization methods to tackle this issue. The most prominent algorithm in this line of research is LARS, which by employing layerwise adaptive learning ra…
Recent work by Brock et al. (2018) suggests that Generative Adversarial Networks (GANs) benefit disproportionately from large mini-batch sizes. Unfortunately, using large batches is slow and expensive on conventional hardware. Thus, it would be nice if we could generate batches that were effectively large though actual…
Large-batch stochastic gradient descent (SGD) is widely used for training in distributed deep learning because of its training-time efficiency, however, extremely large-batch SGD leads to poor generalization and easily converges to sharp minima, which prevents naive large-scale data-parallel SGD (DP-SGD) from convergin…
STORM-PG uses momentum for faster policy gradient updates.
problem Improving policy gradient methods for reinforcement learning.
method Introduces STORM-PG, a SARAH-based algorithm with exponential moving average.
result Achieves sample complexity, matching best-known rate.
Mini-batch stochastic gradient methods (SGD) are state of the art for distributed training of deep neural networks. Drastic increases in the mini-batch sizes have lead to key efficiency and scalability gains in recent years. However, progress faces a major roadblock, as models trained with large batches often do not ge…
ABS dynamically adjusts batch size based on policy stability, improving RL performance.
problem Diminishing returns with large batch sizes in RL due to non-stationary data.
method Adaptive Batch Scaling (ABS) with Behavioral Divergence metric.
result Larger batch sizes can improve RL performance, contrary to conventional wisdom.
Adaptor 'E' extends gradient-based optimizers to explore loss landscapes, improving generalization.
problem Finding lower and better-generalizing minima in deep learning.
method Proposes an adaptor 'E' to extend gradient-based optimizers, encouraging exploration along landscape valleys.
result Adapted optimizers increase test accuracy by an average of 2.5% in large-batch training tasks.
Background: Deep learning models are typically trained using stochastic gradient descent or one of its variants. These methods update the weights using their gradient, estimated from a small fraction of the training data. It has been observed that when using large batch sizes there is a persistent degradation in genera…
SogCLR uses small batch sizes for global contrastive learning, achieving similar performance to SimCLR.
problem Existing contrastive learning methods require large batch sizes or large feature dictionaries.
method SogCLR, a memory-efficient Stochastic Optimization algorithm for global contrastive learning.
result SogCLR with small batch sizes (e.g., 256) achieves similar performance to SimCLR with large batch sizes (e.g., 8192).
New approach speeds up optimization with repeated gradient steps on same batch.
problem Performance bottlenecks in massive parallel pipelines with large batch sizes.
method Data echoing, taking repeated gradient steps on the same batch.
result Data echoing affords speedups on curvature-dominated part of convergence rate.
A One-Sample Decentralized Proximal Algorithm for Non-Convex Stochastic Composite Optimizationmath.OC
Two algorithms find optimal points in decentralized optimization.
problem Decentralized non-convex stochastic optimization with composite objective functions.
method Prox-DASA and Prox-DASA-GT algorithms for finding ε-stationary points.
result Achieves comparable complexity without large batch sizes or complex per-iteration operations.
Stochastic gradient descent (SGD) is almost ubiquitously used for training non-convex optimization tasks. Recently, a hypothesis proposed by Keskar et al. [2017] that large batch methods tend to converge to sharp minimizers has received increasing attention. We theoretically justify this hypothesis by providing new pro…
Training deep neural networks with Stochastic Gradient Descent, or its variants, requires careful choice of both learning rate and batch size. While smaller batch sizes generally converge in fewer training epochs, larger batch sizes offer more parallelism and hence better computational efficiency. We have developed a n…
A new method improves active learning for large batch sizes.
problem Challenges in scaling Bayesian active learning to large batch sizes.
method Derives Partial Batch Label Sampling (ParBaLS) for EPIG algorithm.
result ParBaLS EPIG outperforms top- selection and BatchBALD.
Empirical study on SGD hyperparameters and adversarial robustness.
problem Effect of SGD hyperparameters on adversarial robustness and generalization.
method Empirical observation of learning rate, batch size, and momentum effects on adversarial robustness and generalization.
result Constant learning rate to batch size ratio leads to good generalization and almost constant adversarial robustness.
Sarah Frank-Wolfe: Methods for Constrained Optimization with Best Rates and Practical Featuresmath.OC
Two new Frank-Wolfe algorithms improve convergence for constrained optimization.
problem Solving optimization problems with structured constraints in machine learning.
method Two new variants of the Frank-Wolfe (FW) method for stochastic finite-sum minimization.
result Best convergence guarantees for convex and non-convex objective functions.
Critical Parameters for Scalable Distributed Learning with Large Batches and Asynchronous Updatescs.LG
New insights explain speedup saturation in distributed learning with large batches and delays.
problem Understanding and optimizing speedup in distributed learning with large batches and delays.
method Theoretical analysis of strongly convex, convex, and non-convex settings, considering data sparsity.
result Identification of a data-dependent parameter explaining speedup saturation in both batch size and gradient staleness.
Stochastic Gradient Descent (SGD) and its variants are mainstream methods for training deep networks in practice. SGD is known to find a flat minimum that often generalizes well. However, it is mathematically unclear how deep learning can select a flat minimum among so many minima. To answer the question quantitatively…
SGDM accelerates faster than SGD with large batch sizes and permits broader learning rates.
problem Understanding the role of momentum in SGDM and its convergence rates.
method Analysis of SGDM convergence rates under strongly convex settings, including finite-sample rates and asymptotic normality of the averaged estimator.
result SGDM converges faster than SGD with large batch sizes and permits broader learning rates.
Accurate Large-sample Uncertainty Quantification using Stochastic Gradient Markov Chain Monte Carlocs.LG
New method improves uncertainty quantification for large batch sizes and misspecified models.
problem Challenges in tuning algorithms for accurate uncertainty quantification in large batch sizes and misspecified models.
method Proposes new discrete-time approximations to SGD and SGLD, proving error bounds for practical tuning.
result Quantitative, non-asymptotic error bounds for accurate predictions of covariance and autocorrelation time.
New method improves model accuracy in Byzantine-robust distributed learning by optimizing batch size.
problem Reduces model accuracy drop due to large variance of stochastic gradients in Byzantine-robust distributed learning.
method Proposes ByzSGDnm, a novel BRDL method that uses normalized momentum to mitigate accuracy drop in large batch sizes.
result The optimal batch size increases with the fraction of Byzantine workers, leading to better model accuracy under Byzantine attacks.
Efficient momentum-based methods for reinforcement learning with improved sample complexity.
problem Efficient reinforcement learning algorithms for non-concave performance functions.
method Adaptive momentum-based policy gradient methods using variance reduction and importance sampling techniques.
result Both IS-MBPG and HA-MBPG reach the best known sample complexity of for finding an -stationary point.
In an increasing number of domains it has been demonstrated that deep learning models can be trained using relatively large batch sizes without sacrificing data efficiency. However the limits of this massive data parallelism seem to differ from domain to domain, ranging from batches of tens of thousands in ImageNet to …
Increasing the mini-batch size for stochastic gradient descent offers significant opportunities to reduce wall-clock training time, but there are a variety of theoretical and systems challenges that impede the widespread success of this technique. We investigate these issues, with an emphasis on time to convergence and…
Current state-of-the-art NMT systems use large neural networks that are not only slow to train, but also often require many heuristics and optimization tricks, such as specialized learning rate schedules and large batch sizes. This is undesirable as it requires extensive hyperparameter tuning. In this paper, we propose…