FLORES-200 Kurdish Evaluation Benchmark

Year: 2022
Size: 3,001 designed sentences per language; 2,009 publicly released dev/devtest sentences per Kurdish variety
License: CC BY-SA 4.0
Domain: Machine Translation Evaluation
Languages:
Central Kurdish Northern Kurdish English Multilingual
Formats:
TXT

Resource Description

Multilingual machine translation evaluation benchmark containing Central Kurdish (ckb_Arab) and Northern Kurdish (kmr_Latn). Public Kurdish portions contain 997 development and 1,012 development-test sentences per language; the remaining test portion is hidden.

Data Structure

Multiway-aligned UTF-8 evaluation text with standardized language identifiers.

Primary Publication

Scaling neural machine translation to 200 languages

NLLB Team (2024)

Presents the No Language Left Behind multilingual machine translation system. The released model family explicitly supports Central Kurdish in Arabic script and Northern Kurdish in Latin script among …

Related Publications

How to Cite

NLLB Team. FLORES-200 evaluation benchmark.

Dataset Information

Total Size 3,001 designed sentences per language; 2,009 publicly released dev/devtest sentences per Kurdish variety
Languages 4 languages
Formats 1 format
Related Papers 1 publication

Data Access

Download Dataset

Licensed under CC BY-SA 4.0

View Main Paper Back to Datasets