liu-nlp/blimp-single-error

该数据集是一个多语言数据集，包含五种语言的配置：爱沙尼亚语（et）、波斯语（fa）、法罗语（fo）、冰岛语（is）和瑞典语（sv）。每个配置包含诸如correct和incorrect字符串等特征，以及代表各种语法转换或语言现象的分割。数据集的结构支持与语法错误纠正或语言分析相关的任务，每个分割代表一种特定类型的转换（例如，属格到入格、单数到复数等）。提供了每个配置的数据集大小和下载大小，表明了数据的规模。

This dataset is a multilingual dataset with configurations for five languages: Estonian (et), Persian (fa), Faroese (fo), Icelandic (is), and Swedish (sv). Each configuration includes features such as correct and incorrect strings, and splits that represent various grammatical transformations or linguistic phenomena. The dataset is structured to support tasks related to grammatical error correction or linguistic analysis, with each split representing a specific type of transformation (e.g., genitive to illative, singular to plural, etc.). The dataset sizes and download sizes are provided for each configuration, indicating the scale of the data.