aale1/ag_news_annotated

该数据集名为ag_news_annotated，是一个基于AG News新闻文章数据集的标注版本，使用Argilla平台创建。数据集规模在10万到100万条记录之间，包含文本字段（如新闻内容）和标注任务，涉及文本分类（类别包括商业、科技、体育和世界）以及实体标注（如识别文本中的实体）。数据集旨在通过人工反馈进行标注，适用于自然语言处理任务，如分类和实体识别。数据实例以JSON格式提供，包含原始文本、标注建议和状态信息，数据分割为单一的train集。The dataset named ag_news_annotated is an annotated version of the AG News news article dataset, created using the Argilla platform. It has a size between 100,000 and 1,000,000 records and includes text fields (e.g., news content) and annotation tasks, covering text classification (with categories such as Business, Sci/Tech, Sports, and World) and entity annotation (e.g., identifying entities in the text). The dataset is designed for human feedback annotation and is suitable for natural language processing tasks like classification and entity recognition. Data instances are provided in JSON format, including raw text, annotation suggestions, and status information, with a single train split.