基于多GPU并行框架的DNN语音识别研究  

Speech Recognition Research on Multi-GPU Parallel Framework of Deep Neural Network

作　　者：杨宁[1] 

机构地区：[1]南京晓庄学院数学与信息技术学院,江苏南京211171

出　　处：《微电子学与计算机》2015年第7期6-10,共5页Microelectronics & Computer

基　　金：国家自然科学基金青年基金项目(61202136)

摘　　要：提出了深度神经网络DNN的多GPU并行框架,描述了其实现方法及其性能优化,依托多GPU的强大协同并行计算能力,结合数据并行特点,实现快速高效的深度神经网络训练.对语音识别应用,在模型收敛速度和模型性能上都取得了有效提升——相比单GPU有4.6倍加速比,数十亿样本的训练数天收敛,字错率降低约10%.A Muhi-GPU parallel framework o{ DNN is given out, and the implementation method and its performance optimization are presented, relying on the powerful synergy parallel computing ability of Muhi-GPU, combining with the characteristics of data parallel, the fast and efficient training of Deep Neural Network is realized. The application of speech recognition, the model convergence rate and model performance have achieved effective promotion. Compared with single GPU, it has 4. 6 times speedup ratio, billions of training days convergence, and word error rate is reduced about 10%.

关 键 词：深度神经网络 语音识别 图形处理器 并行框架 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]