Semi-Autoregressive Training Improves Mask-Predict Decoding

Zettlemoyer Luke
Zettlemoyer Luke
Cited by: 5|Bibtex|Views38
Other Links: arxiv.org

Abstract:

The recently proposed mask-predict decoding algorithm has narrowed the performance gap between semi-autoregressive machine translation models and the traditional left-to-right approach. We introduce a new training method for conditional masked language models, SMART, which mimics the semi-autoregressive behavior of mask-predict, produci...More

Code:

Data:

Full Text
Your rating :
0

 

Tags
Comments