arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1802.04680cs.LG
收藏

深度神经网络中整数的训练和推理

Training and Inference with Integers in Deep Neural Networks

Department of Precision Instrument(精密仪器系);Department of Automation(自动化系);Center for Brain Inspired Computing Research(脑启发计算研究中心);Beijing Innovation Center for Future Chip(北京未来芯片创新中心);Tsinghua University(清华大学)

Shuang Wu, Guoqi Li, Feng Chen, Luping Shi

摘要:

具有离散参数的深度神经网络及其在嵌入式系统中的应用研究一直是一个活跃而有前途的课题。虽然以前的工作已经成功地降低了推理的精度,但将训练和推理过程同时转移到低位宽整数还没有被证明。在这项工作中,我们开发了一种称为“WAGE”的新方法来离散化训练和推理,其中层之间的权重(W),激活(A),梯度(G)和错误(E)被移位并线性约束为低位宽整数。为了对定点设备执行纯离散采样,我们进一步用常数缩放层代替批量归一化,并简化了整数实现的其他组件。可以在多个数据集上获得更高的精度,这表明WAGE在某种程度上充当了一种正则化。从经验上讲,我们展示了在硬件系统中部署训练的潜力,例如基于整数的深度学习加速器和神经形态芯片,具有相当的准确性和更高的能效,这对于未来AI应用程序在具有迁移和持续学习需求的可变场景中至关重要。

英文摘要:

Researches on deep neural networks with discrete parameters and their deployment in embedded systems have been active and promising topics. Although previous works have successfully reduced precision in inference, transferring both training and inference processes to low-bitwidth integers has not been demonstrated simultaneously. In this work, we develop a new method termed as "WAGE" to discretize both training and inference, where weights (W), activations (A), gradients (G) and errors (E) among layers are shifted and linearly constrained to low-bitwidth integers. To perform pure discrete dataflow for fixed-point devices, we further replace batch normalization by a constant scaling layer and simplify other components that are arduous for integer implementation. Improved accuracies can be obtained on multiple datasets, which indicates that WAGE somehow acts as a type of regularization. Empirically, we demonstrate the potential to deploy training in hardware systems such as integer-based deep learning accelerators and neuromorphic chips with comparable accuracy and higher energy efficiency, which is crucial to future AI applications in variable scenarios with transfer and continual learning demands.

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。

↑