BriefGPT.xyz
Feb, 2020
从英语到外语:迁移预训练语言模型
From English To Foreign Languages: Transferring Pre-trained Language Models
HTML
PDF
Ke Tran
TL;DR
本文介绍一种在有限计算预算下将英文预训练模型转移到其他语言的方法,使用单个GPU,一天内可以获得一种新的外语 BERT 基础模型,并在六种语言上展示该方法在零样本任务上比多语言 BERT 更为有效的结果。
Abstract
pre-trained models
have demonstrated their effectiveness in many downstream natural language processing (
nlp
) tasks. The availability of
multilin
→