pietrolesci/opus-raw

该数据集包含多个语言对的平行语料，涵盖德语-英语、希腊语-英语、英语-法语、英语-意大利语、英语-立陶宛语、英语-荷兰语、英语-葡萄牙语和英语-西班牙语。每个语言对的配置包括训练集、验证集和测试集，分别用于模型的训练、验证和测试。数据集的特征包括id、源语言和目标语言的文本。每个分割的数据大小和示例数量在README中有详细说明。

This dataset contains parallel corpora for multiple language pairs, including German-English, Greek-English, English-French, English-Italian, English-Lithuanian, English-Dutch, English-Portuguese, and English-Spanish. Each language pair configuration includes training, validation, and test sets, which are used for model training, validation, and testing, respectively. The dataset features include id, source language text, and target language text. The size and number of examples for each split are detailed in the README.