基于知识蒸馏的中文语法错误纠正

Jul, 2022

基于知识蒸馏的中文语法错误纠正

Chinese grammatical error correction based on knowledge distillation

Peng Xia, Yuechi Zhou, Ziyan Zhang, Zecheng Tang, Juntao Li

TL;DR使用知识蒸馏方法压缩模型参数，构建攻击测试集，从而在中文语法错误纠正上提高模型抗攻性和鲁棒性。实验结果表明，压缩小型模型的性能仍然良好，训练速度加快，具有最佳效果和显著的鲁棒性。

Abstract

In view of the poor robustness of existing chinese grammatical error correction models on attack test sets and large model parameters, this paper uses the method of →