sumuks/celerai_education

该数据集包含了多个配置，每个配置根据不同的NLP任务有不同的特征，如文档摘要、多跳问题回答、单次问题回答等。具体特征包括文档ID、文档文本、文档文件名、文档元数据（如文件大小）、摘要模型、问题生成模型、答案、问题类别、难度评估等。数据集主要包含训练集分割，并提供了对应的数据文件路径。

The dataset consists of multiple configurations, each tailored to different NLP tasks with specific features such as document summarization, multi-hop question answering, single-shot question answering, etc. Features include document ID, document text, document filename, document metadata (e.g., file size), summarization model, question generating model, answers, question category, difficulty assessment, etc. The dataset primarily contains a training split and provides corresponding data file paths.