2017/12/08 by Shankar Krishnan, Ying Xiao, Krishnan, Shankar +3
Computer Science · Mathematics · #Advanced Image and Video Retrieval Techniques #Advanced Neural Network Applications #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Data Classification #Stochastic Gradient Optimization Techniques #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.1712.03298
arxiv created 2017/12/08 · openalex publication_date 2017/12/08 · arxiv updated 2017/12/12 · openalex created_date 2022/09/30 · openalex updated_date 2026/07/28
Progress in deep learning is slowed by the days or weeks it takes to train large models. The natural solution of using more hardware is limited by diminishing returns, and leads to inefficient use of additional resources. In this paper, we present a large batch, stochastic optimization algorithm that is both faster than widely used algorithms for fixed amounts of computation, and also scales up substantially better as more computational resources become available. Our algorithm implicitly computes the inverse Hessian of each mini-batch to produce descent directions; we do so without either an explicit approximation to the Hessian or Hessian-vector products. We demonstrate the effectiveness of our algorithm by successfully training large ImageNet models (Inception-V3, Resnet-50, Resnet-101 and Inception-Resnet-V2) with mini-batch sizes of up to 32000 with no loss in validation error relative to current baselines, and no increase in the total number of steps. At smaller mini-batch sizes, our optimizer improves the validation error in these models by 0.8-0.9%. Alternatively, we can trade off this accuracy to reduce the number of training steps needed by roughly 10-30%. Our work is practical and easily usable by others -- only one hyperparameter (learning rate) needs tuning, and furthermore, the algorithm is as computationally cheap as the commonly used Adam optimizer.