TY - GEN
T1 - Neural machine translation with byte-level subwords
AU - Wang, Changhan
AU - Cho, Kyunghyun
AU - Gu, Jiatao
N1 - Publisher Copyright:
Copyright © 2020 Association for the Advancement of Artificial Intelligence. All rights reserved.
PY - 2020
Y1 - 2020
N2 - Almost all existing machine translation models are built on top of character-based vocabularies: characters, subwords or words. Rare characters from noisy text or character-rich languages such as Japanese and Chinese however can unnecessarily take up vocabulary slots and limit its compactness. Representing text at the level of bytes and using the 256 byte set as vocabulary is a potential solution to this issue. High computational cost has however prevented it from being widely deployed or used in practice. In this paper, we investigate byte-level subwords, specifically byte-level BPE (BBPE), which is compacter than character vocabulary and has no out-of-vocabulary tokens, but is more efficient than using pure bytes only is. We claim that contextualizing BBPE embeddings is necessary, which can be implemented by a convolutional or recurrent layer. Our experiments show that BBPE has comparable performance to BPE while its size is only 1/8 of that for BPE. In the multilingual setting, BBPE maximizes vocabulary sharing across many languages and achieves better translation quality. Moreover, we show that BBPE enables transferring models between languages with non-overlapping character sets.
AB - Almost all existing machine translation models are built on top of character-based vocabularies: characters, subwords or words. Rare characters from noisy text or character-rich languages such as Japanese and Chinese however can unnecessarily take up vocabulary slots and limit its compactness. Representing text at the level of bytes and using the 256 byte set as vocabulary is a potential solution to this issue. High computational cost has however prevented it from being widely deployed or used in practice. In this paper, we investigate byte-level subwords, specifically byte-level BPE (BBPE), which is compacter than character vocabulary and has no out-of-vocabulary tokens, but is more efficient than using pure bytes only is. We claim that contextualizing BBPE embeddings is necessary, which can be implemented by a convolutional or recurrent layer. Our experiments show that BBPE has comparable performance to BPE while its size is only 1/8 of that for BPE. In the multilingual setting, BBPE maximizes vocabulary sharing across many languages and achieves better translation quality. Moreover, we show that BBPE enables transferring models between languages with non-overlapping character sets.
UR - http://www.scopus.com/inward/record.url?scp=85098441562&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85098441562&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85098441562
T3 - AAAI 2020 - 34th AAAI Conference on Artificial Intelligence
SP - 9154
EP - 9160
BT - AAAI 2020 - 34th AAAI Conference on Artificial Intelligence
PB - AAAI press
T2 - 34th AAAI Conference on Artificial Intelligence, AAAI 2020
Y2 - 7 February 2020 through 12 February 2020
ER -