mteb/CrossLingualSemanticDiscriminationWMT19
收藏Hugging Face2025-05-06 更新2025-05-31 收录
下载链接：
https://hf-mirror.com/datasets/mteb/CrossLingualSemanticDiscriminationWMT19
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是MTEB（Massive Text Embedding Benchmark）的一部分，用于评估多语言嵌入模型在区分原始平行对与具有挑战性的干扰项方面的能力，这些干扰项源自WMT19德语-法语测试集。数据集包括德语-法语和法语-德语语料库、查询和相关性评分的配置。任务类别包括文本检索和翻译到翻译（t2t）。数据集在CC BY-SA 4.0许可下，支持德语（deu）和法语（fra）。
This dataset is part of the MTEB (Massive Text Embedding Benchmark) and is used to evaluate multilingual embedding models based on their ability to discriminate against the original parallel pair against challenging distractors, derived from the WMT19 DE-FR test set. The dataset includes configurations for German-French and French-German corpora, queries, and relevance scores. The task categories are text-retrieval and translation-to-translation (t2t). The dataset is licensed under CC BY-SA 4.0 and is multilingual, supporting German (deu) and French (fra).
提供机构：
mteb
