arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.14917cs.LGstat.ML
收藏

通过层融合压缩深度神经网络

Compressing Deep Neural Networks via Layer Fusion

University of Liverpool(利物浦大学);USC Information Sciences Institute(南加州大学信息科学研究所)

James O' Neill, Greg Ver Steeg, Aram Galstyan

摘要:

本文提出了\textit{layer fusion} -一种模型压缩技术,该技术发现要组合哪些权重,然后融合相似的全连接层、卷积层和注意力层的权重。层融合可以显著减少原始网络的层数,同时保持有竞争力的性能。从CIFAR-10上的实验中,我们发现,当使用层融合进行迭代再训练时,各种深度卷积神经网络可以保持在原始网络的2%精度点以内,压缩比高达3.33。对于使用预训练的Transformer模型的WikiText-2语言建模数据集的实验,我们实现了压缩,导致网络是其原始大小的20%,同时在原始网络的5个困惑点内。我们还发现,与原始网络相比,其他成熟的压缩技术可以实现具有竞争力的性能,只要有足够数量的再训练步骤。一般来说,我们观察到一个明确的拐点的性能作为压缩量的增加,这表明在性能指数下降之前可以实现的压缩量的约束。

英文摘要:

This paper proposes \textit{layer fusion} - a model compression technique that discovers which weights to combine and then fuses weights of similar fully-connected, convolutional and attention layers. Layer fusion can significantly reduce the number of layers of the original network with little additional computation overhead, while maintaining competitive performance. From experiments on CIFAR-10, we find that various deep convolution neural networks can remain within 2\% accuracy points of the original networks up to a compression ratio of 3.33 when iteratively retrained with layer fusion. For experiments on the WikiText-2 language modelling dataset where pretrained transformer models are used, we achieve compression that leads to a network that is 20\% of its original size while being within 5 perplexity points of the original network. We also find that other well-established compression techniques can achieve competitive performance when compared to their original networks given a sufficient number of retraining steps. Generally, we observe a clear inflection point in performance as the amount of compression increases, suggesting a bound on the amount of compression that can be achieved before an exponential degradation in performance.

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。

↑