The RoyalFlush System for the WMT 2022 Efficiency Task

12/03/2022
by   Bo Qin, et al.
0

This paper describes the submission of the RoyalFlush neural machine translation system for the WMT 2022 translation efficiency task. Unlike the commonly used autoregressive translation system, we adopted a two-stage translation paradigm called Hybrid Regression Translation (HRT) to combine the advantages of autoregressive and non-autoregressive translation. Specifically, HRT first autoregressively generates a discontinuous sequence (e.g., make a prediction every k tokens, k>1) and then fills in all previously skipped tokens at once in a non-autoregressive manner. Thus, we can easily trade off the translation quality and speed by adjusting k. In addition, by integrating other modeling techniques (e.g., sequence-level knowledge distillation and deep-encoder-shallow-decoder layer allocation strategy) and a mass of engineering efforts, HRT improves 80% inference speed and achieves equivalent translation performance with the same-capacity AT counterpart. Our fastest system reaches 6k+ words/second on the GPU latency setting, estimated to be about 3.1x faster than the last year's winner.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/19/2022

Hybrid-Regressive Neural Machine Translation

In this work, we empirically confirm that non-autoregressive translation...
research
06/18/2020

Deep Encoder, Shallow Decoder: Reevaluating the Speed-Quality Tradeoff in Machine Translation

State-of-the-art neural machine translation models generate outputs auto...
research
09/16/2021

The NiuTrans System for the WMT21 Efficiency Task

This paper describes the NiuTrans system for the WMT21 translation effic...
research
12/31/2020

Fully Non-autoregressive Neural Machine Translation: Tricks of the Trade

Fully non-autoregressive neural machine translation (NAT) is proposed to...
research
10/14/2021

Non-Autoregressive Translation with Layer-Wise Prediction and Deep Supervision

How do we perform efficient inference while retaining high translation q...
research
06/05/2022

Multilingual Neural Machine Translation with Deep Encoder and Multiple Shallow Decoders

Recent work in multilingual translation advances translation quality sur...
research
07/17/2020

Task-Level Curriculum Learning for Non-Autoregressive Neural Machine Translation

Non-autoregressive translation (NAT) achieves faster inference speed but...

Please sign up or login with your details

Forgot password? Click here to reset