BriefGPT.xyz
Dec, 2020
从阅读器向检索器提取知识以作问答之用
Distilling Knowledge from Reader to Retriever for Question Answering
HTML
PDF
Gautier Izacard, Edouard Grave
TL;DR
本文提出了一种受到知识蒸馏启发的学习信息检索器的技术,在没有标注的查询和文档对的情况下,利用阅读器模型的注意力分数来获得检索器的合成标签,用于下游任务,最终在问题回答方面达到了最先进的结果。
Abstract
The task of
information retrieval
is an important component of many natural language processing systems, such as open domain
question answering
. While traditional methods were based on hand-crafted features, cont
→