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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

We propose a two-stage approach for training a single NMT model to translateunseen languages both to and from English. For the first stage, we initializean encoder-decoder model to pretrained XLM-R and RoBERTa weights, then performmultilingual fine-tuning on parallel data in 40 languages to English. We findthis model can generalize to zero-shot translations on unseen languages. Forthe second stage, we leverage this generalization ability to generate syntheticparallel data from monolingual datasets, then bidirectionally train withsuccessive rounds of back-translation.  Our approach, which we EcXTra (English-centric Crosslingual (X) Transfer), isconceptually simple, only using a standard cross-entropy objective throughout.It is also data-driven, sequentially leveraging auxiliary parallel data andmonolingual data. We evaluate unsupervised NMT results for 7 low-resourcelanguages, and find that each round of back-translation training furtherrefines bidirectional performance. Our final single EcXTra-trained modelachieves competitive translation performance in all translation directions,notably establishing a new state-of-the-art for English-to-Kazakh (22.9 > 10.4BLEU). Our code is available at https://github.com/manestay/EcXTra .

Multilingual Bidirectional Unsupervised Translation Through Multilingual Finetuning and Back-Translation