English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recently, non-autoregressive (NAR) neural machine translation models havereceived increasing attention due to their efficient parallel decoding.However, the probabilistic framework of NAR models necessitates conditionalindependence assumption on target sequences, falling short of characterizinghuman language data. This drawback results in less informative learning signalsfor NAR models under conventional MLE training, thereby yielding unsatisfactoryaccuracy compared to their autoregressive (AR) counterparts. In this paper, wepropose a simple and model-agnostic multi-task learning framework to providemore informative learning signals. During training stage, we introduce a set ofsufficiently weak AR decoders that solely rely on the information provided byNAR decoder to make prediction, forcing the NAR decoder to become stronger orelse it will be unable to support its weak AR partners. Experiments on WMT andIWSLT datasets show that our approach can consistently improve accuracy ofmultiple NAR baselines without adding any additional decoding overhead.

Helping the Weak Makes You Strong: Simple Multi-Task Learning Improves Non-Autoregressive Translators