开放域问答中可扩展推理的解耦变压器

Haytham ElFadeel, Stanislav Peshterliev
{"title":"开放域问答中可扩展推理的解耦变压器","authors":"Haytham ElFadeel, Stanislav Peshterliev","doi":"10.26615/978-954-452-072-4_044","DOIUrl":null,"url":null,"abstract":"Large transformer models, such as BERT, achieve state-of-the-art results in machine reading comprehension (MRC) for open-domain question answering (QA). However, transformers have a high computational cost for inference which makes them hard to apply to online QA systems for applications like voice assistants. To reduce computational cost and latency, we propose decoupling the transformer MRC model into input-component and cross-component. The decoupling allows for part of the representation computation to be performed offline and cached for online use. To retain the decoupled transformer accuracy, we devised a knowledge distillation objective from a standard transformer model. Moreover, we introduce learned representation compression layers which help reduce by four times the storage requirement for the cache. In experiments on the SQUAD 2.0 dataset, a decoupled transformer reduces the computational cost and latency of open-domain MRC by 30-40% with only 1.2 points worse F1-score compared to a standard transformer.","PeriodicalId":284493,"journal":{"name":"Recent Advances in Natural Language Processing","volume":"23 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Decoupled Transformer for Scalable Inference in Open-domain Question Answering\",\"authors\":\"Haytham ElFadeel, Stanislav Peshterliev\",\"doi\":\"10.26615/978-954-452-072-4_044\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Large transformer models, such as BERT, achieve state-of-the-art results in machine reading comprehension (MRC) for open-domain question answering (QA). However, transformers have a high computational cost for inference which makes them hard to apply to online QA systems for applications like voice assistants. To reduce computational cost and latency, we propose decoupling the transformer MRC model into input-component and cross-component. The decoupling allows for part of the representation computation to be performed offline and cached for online use. To retain the decoupled transformer accuracy, we devised a knowledge distillation objective from a standard transformer model. Moreover, we introduce learned representation compression layers which help reduce by four times the storage requirement for the cache. In experiments on the SQUAD 2.0 dataset, a decoupled transformer reduces the computational cost and latency of open-domain MRC by 30-40% with only 1.2 points worse F1-score compared to a standard transformer.\",\"PeriodicalId\":284493,\"journal\":{\"name\":\"Recent Advances in Natural Language Processing\",\"volume\":\"23 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2021-08-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Recent Advances in Natural Language Processing\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.26615/978-954-452-072-4_044\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Recent Advances in Natural Language Processing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.26615/978-954-452-072-4_044","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1

摘要

大型变压器模型,如BERT,在开放域问答(QA)的机器阅读理解(MRC)中取得了最先进的结果。然而,变压器的推理计算成本很高,这使得它们很难应用于语音助手等应用程序的在线QA系统。为了减少计算成本和延迟,我们提出将变压器MRC模型解耦为输入组件和跨组件。这种解耦允许脱机执行部分表示计算,并缓存以供在线使用。为了保持解耦变压器的精度,我们设计了一个标准变压器模型的知识蒸馏目标。此外,我们引入了学习表示压缩层,这有助于将缓存的存储需求减少四倍。在SQUAD 2.0数据集上的实验中,与标准变压器相比,解耦变压器将开放域MRC的计算成本和延迟降低了30-40%,f1分数仅差1.2分。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Decoupled Transformer for Scalable Inference in Open-domain Question Answering
Large transformer models, such as BERT, achieve state-of-the-art results in machine reading comprehension (MRC) for open-domain question answering (QA). However, transformers have a high computational cost for inference which makes them hard to apply to online QA systems for applications like voice assistants. To reduce computational cost and latency, we propose decoupling the transformer MRC model into input-component and cross-component. The decoupling allows for part of the representation computation to be performed offline and cached for online use. To retain the decoupled transformer accuracy, we devised a knowledge distillation objective from a standard transformer model. Moreover, we introduce learned representation compression layers which help reduce by four times the storage requirement for the cache. In experiments on the SQUAD 2.0 dataset, a decoupled transformer reduces the computational cost and latency of open-domain MRC by 30-40% with only 1.2 points worse F1-score compared to a standard transformer.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Exploring the Landscape of Natural Language Processing Research AspectCSE: Sentence Embeddings for Aspect-Based Semantic Textual Similarity Using Contrastive Learning and Structured Knowledge Efficient Domain Adaptation of Sentence Embeddings Using Adapters Neural Machine Translation for Sinhala-English Code-Mixed Text A Domain-Independent Holistic Approach to Deception Detection
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1