Self-Paced Learning for Neural Machine Translation

Yu Wan, Baosong Yang, Derek F. Wong, Yikai Zhou, Lidia S. Chao, Haibo Zhang, Boxing Chen


Abstract
Recent studies have proven that the training of neural machine translation (NMT) can be facilitated by mimicking the learning process of humans. Nevertheless, achievements of such kind of curriculum learning rely on the quality of artificial schedule drawn up with the handcrafted features, e.g. sentence length or word rarity. We ameliorate this procedure with a more flexible manner by proposing self-paced learning, where NMT model is allowed to 1) automatically quantify the learning confidence over training examples; and 2) flexibly govern its learning via regulating the loss in each iteration step. Experimental results over multiple translation tasks demonstrate that the proposed model yields better performance than strong baselines and those models trained with human-designed curricula on both translation quality and convergence speed.
Anthology ID:
2020.emnlp-main.80
Volume:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Month:
November
Year:
2020
Address:
Online
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1074–1080
Language:
URL:
https://aclanthology.org/2020.emnlp-main.80
DOI:
10.18653/v1/2020.emnlp-main.80
Bibkey:
Copy Citation:
PDF:
https://aclanthology.org/2020.emnlp-main.80.pdf
Video:
 https://slideslive.com/38938638