arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2101.00542cs.CL
收藏

一种具有压缩子层的高效Transformer解码器

An Efficient Transformer Decoder with Compressed Sub-layers

Northeastern University(东北大学);NiuTrans Research(NiuTrans研究院)

Yanyang Li, Ye Lin, Tong Xiao, Jingbo Zhu

AI总结:

本文提出压缩注意力网络,通过将Transformer解码器的三个子层压缩为一个,在14个WMT任务上实现1.42倍加速且性能与强基线相当,而该强基线本身已比标准基线快2倍。

摘要:

近年来,基于注意力机制的大型编码器-解码器网络(Transformer)因其有效性而变得普遍。但其解码器的高计算复杂度带来了效率低下的问题。通过检查解码器的数学公式,我们表明在温和条件下,可以通过压缩其子层(Transformer的基本构建块)来简化架构,并实现更高的并行度。因此,我们提出了压缩注意力网络,其解码器层仅包含一个子层,而不是三个。在14个WMT机器翻译任务上的大量实验表明,我们的模型速度提高了1.42倍,性能与强基线相当。该强基线已经比广泛使用的标准基线快2倍,且性能无损失。

英文摘要:

The large attention-based encoder-decoder network (Transformer) has become prevailing recently due to its effectiveness. But the high computation complexity of its decoder raises the inefficiency issue. By examining the mathematic formulation of the decoder, we show that under some mild conditions, the architecture could be simplified by compressing its sub-layers, the basic building block of Transformer, and achieves a higher parallelism. We thereby propose Compressed Attention Network, whose decoder layer consists of only one sub-layer instead of three. Extensive experiments on 14 WMT machine translation tasks show that our model is 1.42x faster with performance on par with a strong baseline. This strong baseline is already 2x faster than the widely used standard baseline without loss 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。

↑