arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.09398cs.LG
收藏

Larq Compute Engine:设计、基准测试和部署最先进的二值化神经网络

Larq Compute Engine: Design, Benchmark, and Deploy State-of-the-Art Binarized Neural Networks

Plumerai Research(Plumerai研究院)

Tom Bannink, Arash Bakhtiari, Adam Hillier, Lukas Geiger, Tim de Bruin, Leon Overweel, Jelmer Neeven, Koen Helwegen

摘要:

我们介绍了Larq计算引擎,世界上最快的二进制神经网络(BNN)推理引擎,并使用这个框架来研究有关BNN效率的几个重要问题,并设计了一个新的国家的最先进的BNN架构。LCE提供了高度优化的二进制运算实现,与Pixel 1手机上的全精度卷积相比,将二进制卷积加速了8.5 - 18.5倍。LCE与Larq的集成以及一个复杂的基于MLIR的转换器使用户能够从培训顺利过渡到部署。通过扩展TensorFlow和TensorFlow Lite,LCE支持结合二进制和全精度层的模型,并且可以轻松集成到现有应用程序中。使用LCE,我们分析了现有BNN计算机视觉架构的性能,并开发了QuickNet,这是一种简单,易于重现的BNN,在ImageNet上的延迟和准确性方面优于现有的二进制网络。此外,我们还研究了全精度快捷方式的影响以及MAC数量与模型延迟之间的关系。我们相信,经验性能应该驱动BNN架构设计,并希望这项工作将有助于其他人设计,基准测试和部署二进制模型。

英文摘要:

We introduce Larq Compute Engine, the world's fastest Binarized Neural Network (BNN) inference engine, and use this framework to investigate several important questions about the efficiency of BNNs and to design a new state-of-the-art BNN architecture. LCE provides highly optimized implementations of binary operations and accelerates binary convolutions by 8.5 - 18.5x compared to their full-precision counterparts on Pixel 1 phones. LCE's integration with Larq and a sophisticated MLIR-based converter allow users to move smoothly from training to deployment. By extending TensorFlow and TensorFlow Lite, LCE supports models which combine binary and full-precision layers, and can be easily integrated into existing applications. Using LCE, we analyze the performance of existing BNN computer vision architectures and develop QuickNet, a simple, easy-to-reproduce BNN that outperforms existing binary networks in terms of latency and accuracy on ImageNet. Furthermore, we investigate the impact of full-precision shortcuts and the relationship between number of MACs and model latency. We are convinced that empirical performance should drive BNN architecture design and hope this work will facilitate others to design, benchmark and deploy binary models.

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。

↑