alexisbrooker/yourbench_airportexample

该数据集包含多个配置，每个配置具有不同的特征和训练数据大小。具体包括：chunked（文档分割）、ingested（文档摄入）、lighteval（轻量级评估）、multi_hop_questions（多跳问题）、single_shot_questions（单次问答）和summarized（文档摘要）。每个配置都有文档ID、文本内容、文件名等字段，以及特定于配置的字段，如chunk_info_metrics（片段信息指标）、question（问题）、answer（答案）等。数据集适用于文档处理、文本摘要、问答系统等NLP任务。

The dataset comprises multiple configurations, each with different features and training data sizes. It includes chunked (document chunking), ingested (document ingestion), lighteval (lightweight evaluation), multi_hop_questions (multi-hop questions), single_shot_questions (single-shot Q&A), and summarized (document summarization). Each configuration has fields like document ID, text content, filename, and configuration-specific fields such as chunk_info_metrics (chunk information metrics), question (question), answer (answer), etc. The dataset is suitable for NLP tasks such as document processing, text summarization, and question-answering systems.