pietrolesci/opus-5langs-1M

该数据集包含多个语言对的翻译数据，如英语-德语（eng-deu）、英语-法语（eng-fra）、英语-意大利语（eng-ita）、英语-荷兰语（eng-nld）和英语-西班牙语（eng-spa）。每个语言对的数据集包含100万条示例，特征包括源文本（source）、目标文本（target）、训练和测试似然度、反事实记忆、文本长度、编辑距离、词汇重叠等统计信息。数据集主要用于机器翻译任务，提供了丰富的特征以支持模型训练和评估。

This dataset contains machine translation data for multiple language pairs, including English to German, French, Italian, Dutch, and Spanish translations. Each language pair dataset includes source and target language texts, along with various statistical features and metrics related to translation quality, such as train and test likelihoods, edit distances, word overlaps, etc. The dataset is suitable for training and evaluating machine translation models.