5 datasets found

The dataset used in the paper is a multilingual neural machine translation dataset, consisting of parallel corpora for Italian-Romanian, English-Romanian, English-German, and...

Translation Given Non-Autoregressive a source sentence x, an AT model generates each target word yt conditioned on previously generated ones y<t, leading to high latency on...

Two widely-used resource-rich benchmarks, WMT17 English-Chinese (20M) and WMT19 English-German (36M) translation tasks

The Multi30k dataset is an extension of the Flickr30k dataset, containing 29,000 train images, 1,014 validation images and 1,000 test images. Each image is accompanied with six...

The WMT 2016 English-German dataset is used for testing the proposed approach.