Xavier1234/Asclepius-Synthetic-Clinical-Notes

该数据集名为Asclepius: Synthetic Clincal Notes & Instruction Dataset，主要用于构建临床大型语言模型。数据集采用临床笔记-问题-答案的格式，包含157k个合成的出院摘要指令-答案对。这些数据是通过GPT-3.5从PMC-Patients病例报告中合成的笔记生成的。数据集支持8种任务，包括命名实体识别、缩写扩展、关系提取、时间信息提取、共指消解、释义、摘要和问答。数据集的语言为英语，结构包含synthetic.csv文件，数据字段包括patient_id（PMC-Patients的唯一病例报告ID）、patient（病例报告文本）、question（GPT-3.5根据患者生成的指令）、answer（GPT-3.5根据给定病例报告和问题生成的答案）和task（问题的对应类别）。

This dataset is named Asclepius: Synthetic Clincal Notes & Instruction Dataset and is primarily used for building clinical large language models. The dataset is composed in a Clinical Note - Question - Answer format, containing 157k synthetic discharge summary instruction-answer pairs. These data were generated by synthesizing notes from PMC-Patients case reports using GPT-3.5. The dataset supports 8 tasks, including Named Entity Recognition, Abbreviation Expansion, Relation Extraction, Temporal Information Extraction, Coreference Resolution, Paraphrasing, Summarization, and Question Answering. The dataset is in English and structured with a synthetic.csv file, with data fields including patient_id (unique case report ID from PMC-Patients), patient (case report text), question (GPT-3.5 generated instruction from patient), answer (GPT-3.5 generated answer for given case report and question), and task (corresponding category of the question).