RabidUmarell/coref-combined

该数据集包含文本数据，每个文本都有一个唯一标识符(id)，文本内容(text)以及核心ference链(coref_chains)。文本被进一步分解为句子，每个句子包含句子id、文本、分词、词性标注、词形、语法特征、头词和依赖关系。数据集分为训练集、测试集和验证集，各自有不同的示例数量和大小。提供了默认配置文件，指定了不同split的数据文件路径。

The dataset consists of text entries, each with a unique identifier (id), text content (text), and coreference chains (coref_chains). The text is further broken down into sentences, each containing sentence id, text, tokens, part-of-speech tags, lemmas, morphological features, head words, and dependency relations. The dataset is split into training, test, and validation sets, each with different numbers of examples and sizes. A default configuration file is provided, specifying the paths to data files for different splits.