CUNI-Bergamot Submission at WMT22 General Task

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3A10457006" target="_blank" >RIV/00216208:11320/22:10457006 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.statmt.org/wmt22/pdf/2022.wmt-1.21.pdf" target="_blank" >https://www.statmt.org/wmt22/pdf/2022.wmt-1.21.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

CUNI-Bergamot Submission at WMT22 General Task

Popis výsledku v původním jazyce

We present CUNI-Bergamot submission for WMT22 General translation task. We compete in English → Czech direction. Our submission further explores block backtranslation techniques. In addition to the previous work, we measure performance in terms of COMET score and named entities translation accuracy. We evaluate performance of MBR decoding compared to traditional mixed backtranslation training and we show possible synergy when using both of the techniques simultaneously. The results show that both approaches are effective means of improving translation quality and they yield even better results when combined.

Název v anglickém jazyce

CUNI-Bergamot Submission at WMT22 General Task

Popis výsledku anglicky

We present CUNI-Bergamot submission for WMT22 General translation task. We compete in English → Czech direction. Our submission further explores block backtranslation techniques. In addition to the previous work, we measure performance in terms of COMET score and named entities translation accuracy. We evaluate performance of MBR decoding compared to traditional mixed backtranslation training and we show possible synergy when using both of the techniques simultaneously. The results show that both approaches are effective means of improving translation quality and they yield even better results when combined.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GX19-26934X" target="_blank" >GX19-26934X: Neuronové reprezentace v multimodálním a mnohojazyčném modelování</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů