arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1709.07417cs.AIcs.LGstat.ML
收藏

基于强化学习的神经优化器搜索

Neural Optimizer Search with Reinforcement Learning

Google Brain(谷歌大脑)

Irwan Bello, Barret Zoph, Vijay Vasudevan, Quoc V. Le

摘要:

我们提出了一种方法来自动化发现优化方法的过程,重点是深度学习架构。我们训练一个递归神经网络控制器生成一个字符串在一个特定领域的语言,描述了一个数学更新方程的基础上的原始函数列表,如梯度,梯度的运行平均等控制器的训练与强化学习,以最大限度地提高性能的模型后几个时期。在CIFAR-10上,我们的方法发现了几个更新规则,这些规则比许多常用的优化器更好,例如Adam,RMSProp或SGD,在ConvNet模型上使用和不使用动量。我们引入了两个新的优化器,名为PowerSign和AddSign,我们显示它们可以很好地传输并改进各种不同任务和架构的训练,包括ImageNet分类和Google的神经机器翻译系统。

英文摘要:

We present an approach to automate the process of discovering optimization methods, with a focus on deep learning architectures. We train a Recurrent Neural Network controller to generate a string in a domain specific language that describes a mathematical update equation based on a list of primitive functions, such as the gradient, running average of the gradient, etc. The controller is trained with Reinforcement Learning to maximize the performance of a model after a few epochs. On CIFAR-10, our method discovers several update rules that are better than many commonly used optimizers, such as Adam, RMSProp, or SGD with and without Momentum on a ConvNet model. We introduce two new optimizers, named PowerSign and AddSign, which we show transfer well and improve training on a variety of different tasks and architectures, including ImageNet classification and Google's neural machine translation system.

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。

↑