The dataset NTREX_bn_sd_benchmark provides news test references for Machine Translation (MT) evaluation, focusing on translations from Bengali to Sindhi. It is part of a comprehensive collection supporting translations into 128 target languages and includes document-level information, making it a valuable tool for multilingual MT benchmarking. Designed for the news domain, this dataset facilitates the assessment of translation quality and aids in the development of robust translation systems. Submitted by Microsoft, this resource is essential for researchers and developers working on Bengali-to-Sindhi translation tasks.
© 2026 - Copyright AIKosh. All rights reserved.