English
Related papers

Related papers: A Kronecker-factored approximate Fisher matrix for…

200 papers

Many hardware proposals have aimed to accelerate inference in AI workloads. Less attention has been paid to hardware acceleration of training, despite the enormous societal impact of rapid training of AI models. Physics-based computers,…

Second-order methods such as KFAC can be useful for neural net training. However, they are often memory-inefficient since their preconditioning Kronecker factors are dense, and numerically unstable in low precision as they require matrix…

Machine Learning · Computer Science 2024-07-24 Wu Lin , Felix Dangel , Runa Eschenhagen , Kirill Neklyudov , Agustinus Kristiadi , Richard E. Turner , Alireza Makhzani

In this paper we propose and study a technique to reduce the number of parameters and computation time in fully-connected layers of neural networks using Kronecker product, at a mild cost of the prediction quality. The technique proceeds by…

Neural and Evolutionary Computing · Computer Science 2015-07-23 Shuchang Zhou , Jia-Nan Wu

K-FAC is a successful tractable implementation of Natural Gradient for Deep Learning, which nevertheless suffers from the requirement to compute the inverse of the Kronecker factors (through an eigen-decomposition). This can be very…

Machine Learning · Computer Science 2022-11-28 Constantin Octavian Puiu

We propose a quadratic penalty method for continual learning of neural networks that contain batch normalization (BN) layers. The Hessian of a loss function represents the curvature of the quadratic penalty function, and a…

Machine Learning · Computer Science 2020-04-17 Janghyeon Lee , Hyeong Gwon Hong , Donggyu Joo , Junmo Kim

Natural gradient descent has proven effective at mitigating the effects of pathological curvature in neural network optimization, but little is known theoretically about its convergence properties, especially for \emph{nonlinear} networks.…

Machine Learning · Statistics 2019-10-29 Guodong Zhang , James Martens , Roger Grosse

The second-order optimization methods, notably the D-KFAC (Distributed Kronecker Factored Approximate Curvature) algorithms, have gained traction on accelerating deep neural network (DNN) training on GPU clusters. However, existing D-KFAC…

Machine Learning · Computer Science 2022-07-01 Lin Zhang , Shaohuai Shi , Wei Wang , Bo Li

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…

Machine Learning · Computer Science 2021-04-21 Linjian Ma , Gabe Montague , Jiayu Ye , Zhewei Yao , Amir Gholami , Kurt Keutzer , Michael W. Mahoney

Despite the predominant use of first-order methods for training deep learning models, second-order methods, and in particular, natural gradient methods, remain of interest because of their potential for accelerating training through the use…

Machine Learning · Computer Science 2021-12-23 Yi Ren , Donald Goldfarb

Natural gradient descent (NGD) is a powerful optimization technique for machine learning, but the computational complexity of the inverse Fisher information matrix limits its application in training deep neural networks. To overcome this…

Machine Learning · Computer Science 2024-12-11 Weihua Liu , Said Boumaraf , Jianwu Li , Chaochao Lin , Xiabi Liu , Lijuan Niu , Naoufel Werghi

Natural Gradient Descent (NGD) helps to accelerate the convergence of gradient descent dynamics, but it requires approximations in large-scale deep neural networks because of its high computational cost. Empirical studies have confirmed…

Machine Learning · Statistics 2022-01-12 Ryo Karakida , Kazuki Osawa

Feed-forward neural networks can be understood as a combination of an intermediate representation and a linear hypothesis. While most previous works aim to diversify the representations, we explore the complementary direction by performing…

Machine Learning · Computer Science 2019-10-24 Han Zhao , Yao-Hung Hubert Tsai , Ruslan Salakhutdinov , Geoffrey J. Gordon

In this paper, we propose and study a technique to reduce the number of parameters and computation time in convolutional neural networks. We use Kronecker product to exploit the local structures within convolution and fully-connected…

Computer Vision and Pattern Recognition · Computer Science 2016-02-05 Shuchang Zhou , Jia-Nan Wu , Yuxin Wu , Xinyu Zhou

This work proposes a time-efficient Natural Gradient Descent method, called TENGraD, with linear convergence guarantees. Computing the inverse of the neural network's Fisher information matrix is expensive in NGD because the Fisher matrix…

Machine Learning · Computer Science 2022-03-04 Saeed Soori , Bugra Can , Baourun Mu , Mert Gürbüzbalaban , Maryam Mehri Dehnavi

Online Continual Learning (OCL) for image classification represents a challenging subset of Continual Learning, focusing on classifying images from a stream without assuming data independence and identical distribution (i.i.d). The primary…

Machine Learning · Computer Science 2026-03-24 Joe Khawand , David Colliaux

A key challenge for gradient based optimization methods in model-free reinforcement learning is to develop an approach that is sample efficient and has low variance. In this work, we apply Kronecker-factored curvature estimation technique…

Machine Learning · Computer Science 2018-12-12 Mohammad Firouzi

Natural Gradient Descent, a second-degree optimization method motivated by the information geometry, makes use of the Fisher Information Matrix instead of the Hessian which is typically used. However, in many cases, the Fisher Information…

Machine Learning · Computer Science 2023-03-10 Rajesh Shrestha

Deep neural networks (DNNs) are currently predominantly trained using first-order methods. Some of these methods (e.g., Adam, AdaGrad, and RMSprop, and their variants) incorporate a small amount of curvature information by using a diagonal…

Machine Learning · Computer Science 2022-10-28 Achraf Bahamou , Donald Goldfarb , Yi Ren

Existing methods of pruning deep neural networks focus on removing unnecessary parameters of the trained network and fine tuning the model afterwards to find a good solution that recovers the initial performance of the trained model. Unlike…

Machine Learning · Computer Science 2021-11-17 Abdolghani Ebrahimi , Diego Klabjan

Feature whitening is a known technique for speeding up training of DNN. Under certain assumptions, whitening the activations reduces the Fisher information matrix to a simple identity matrix, in which case stochastic gradient descent is…

Machine Learning · Computer Science 2020-11-10 Shmulik Markovich-Golan , Barak Battash , Amit Bleiweiss