arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.08877cs.LGmath.OCstat.ML
收藏

训练深度神经网络的实用拟牛顿方法

Practical Quasi-Newton Methods for Training Deep Neural Networks

Columbia University(哥伦比亚大学)

Donald Goldfarb, Yi Ren, Achraf Bahamou

摘要:

我们考虑了实际随机拟牛顿的发展,特别是Kronecker因子块对角BFGS和L-BFGS方法,用于训练深度神经网络(DNN)。在DNN训练中,梯度$n$的变量和分量的数量通常是数千万的量级,并且Hessian具有$n^2$元素。因此,计算和存储完整的$n \times n$ BFGS近似或存储适度数量的(步长、梯度变化)向量对以用于L-BFGS实现是不可能的。在我们提出的方法中,我们通过块对角矩阵来近似Hessian,并使用梯度和Hessian的结构来进一步近似这些块,每个块对应于一个层,作为两个小得多的矩阵的Kronecker积。这类似于KFAC中的方法,该方法在随机自然梯度方法中计算Fisher矩阵的Kronecker因子块对角近似。由于DNN中Hessian的不确定性和高度可变性,我们还提出了一种新的阻尼方法来保持BFGS和L-BFGS近似的上界和下界有界。在对应用于三个数据集的具有九层或十三层的自动编码器前馈神经网络模型的测试中,我们的方法优于或优于KFAC和最先进的一阶随机方法。

英文摘要:

We consider the development of practical stochastic quasi-Newton, and in particular Kronecker-factored block-diagonal BFGS and L-BFGS methods, for training deep neural networks (DNNs). In DNN training, the number of variables and components of the gradient $n$ is often of the order of tens of millions and the Hessian has $n^2$ elements. Consequently, computing and storing a full $n \times n$ BFGS approximation or storing a modest number of (step, change in gradient) vector pairs for use in an L-BFGS implementation is out of the question. In our proposed methods, we approximate the Hessian by a block-diagonal matrix and use the structure of the gradient and Hessian to further approximate these blocks, each of which corresponds to a layer, as the Kronecker product of two much smaller matrices. This is analogous to the approach in KFAC, which computes a Kronecker-factored block-diagonal approximation to the Fisher matrix in a stochastic natural gradient method. Because the indefinite and highly variable nature of the Hessian in a DNN, we also propose a new damping approach to keep the upper as well as the lower bounds of the BFGS and L-BFGS approximations bounded. In tests on autoencoder feed-forward neural network models with either nine or thirteen layers applied to three datasets, our methods outperformed or performed comparably to KFAC and state-of-the-art first-order stochastic methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑