arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1712.02427cs.LG
收藏

移动设备上的高性能超低精度卷积

High performance ultra-low-precision convolutions on mobile devices

Facebook

Andrew Tulloch, Yangqing Jia

摘要:

移动深度学习的许多应用,特别是实时计算机视觉工作负载,都受到计算能力的限制。对于在较旧的消费类手机上运行的工作负载来说尤其如此,其中典型的设备可能由单核或双核ARMv7 CPU供电。我们提供了一个开源实现和对ARMv7设备上现代深度学习工作负载所需的核心原语的最先进的超低精度(<4位精度)实现的全面分析,并展示了比我们额外的最先进的float 32和int 8基线提高4倍-20倍的加速比。

英文摘要:

Many applications of mobile deep learning, especially real-time computer vision workloads, are constrained by computation power. This is particularly true for workloads running on older consumer phones, where a typical device might be powered by a single- or dual-core ARMv7 CPU. We provide an open-source implementation and a comprehensive analysis of (to our knowledge) the state of the art ultra-low-precision (<4 bit precision) implementation of the core primitives required for modern deep learning workloads on ARMv7 devices, and demonstrate speedups of 4x-20x over our additional state-of-the-art float32 and int8 baselines.

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。

↑