arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.05767cs.CL
收藏

用于高效且准确排序任务的集成Transformer:在问答系统中的应用

Ensemble Transformer for Efficient and Accurate Ranking Tasks: an Application to Question Answering Systems

University of California, Irvine(加利福尼亚大学欧文分校);Allen Institute for AI(艾伦人工智能研究所);Amazon Alexa AI(亚马逊Alexa人工智能部门)

Yoshitomo Matsubara, Luca Soldaini, Eric Lind, Alessandro Moschitti

AI总结:

本文提出CERBERUS多头学生架构,通过将多个异构大型Transformer分别蒸馏到不同排序头中,在保持集成多样性的同时以少量额外参数实现高效准确的答案句选择,性能媲美参数多2.7倍的大模型。

摘要:

大型Transformer模型能够极大提升答案句选择(AS2)任务的表现,但其高昂的计算成本阻碍了它们在许多实际应用中的使用。本文探讨了以下研究问题:如何在不显著增加模型复杂度的情况下,使AS2模型更加准确?为解决这一问题,我们提出了一种多头学生架构(命名为CERBERUS),这是一种高效的神经网络,旨在将大型Transformer的集成模型蒸馏到单个更小的模型中。CERBERUS由两个组件组成:一组用于编码输入的Transformer层堆栈,以及一组排序头;与传统的蒸馏技术不同,每个排序头通过蒸馏不同的大型Transformer架构进行训练,从而保持集成成员的多样性。所得模型仅使用少量额外参数即可捕获异构Transformer模型的知识。我们在三个英文AS2数据集上展示了CERBERUS的有效性;我们提出的方法优于我们考虑的所有单模型蒸馏方法,可与参数多2.7倍且运行慢2.5倍的最先进大型AS2模型相媲美。我们的模型代码可在 https://github.com/amazon-research/wqa-cerberus 获取。

英文摘要:

Large transformer models can highly improve Answer Sentence Selection (AS2) tasks, but their high computational costs prevent their use in many real-world applications. In this paper, we explore the following research question: How can we make the AS2 models more accurate without significantly increasing their model complexity? To address the question, we propose a Multiple Heads Student architecture (named CERBERUS), an efficient neural network designed to distill an ensemble of large transformers into a single smaller model. CERBERUS consists of two components: a stack of transformer layers that is used to encode inputs, and a set of ranking heads; unlike traditional distillation technique, each of them is trained by distilling a different large transformer architecture in a way that preserves the diversity of the ensemble members. The resulting model captures the knowledge of heterogeneous transformer models by using just a few extra parameters. We show the effectiveness of CERBERUS on three English datasets for AS2; our proposed approach outperforms all single-model distillations we consider, rivaling the state-of-the-art large AS2 models that have 2.7x more parameters and run 2.5x slower. Code for our model is available at https://github.com/amazon-research/wqa-cerberus

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。

↑