HPAI-BSC/Polymed-QA

Polymed-QA数据集是一个为了提高自动诊断系统性能而开发的合成问答对数据集。它包含了医学知识图谱数据和诊断案例数据，旨在提供全面评估、多样化的疾病信息、有效利用外部知识，并接近现实世界的任务。数据集由Llama-3.1-70B-Instruct模型生成，问题的描述涉及患者的症状和背景，答案基于真实情况提供正确的诊断。

Polymed-QA is a synthetic question-answering dataset developed to improve the performance of Automatic Diagnosis Systems (ADS). It incorporates medical knowledge graph data and diagnosis case data, aiming to provide comprehensive evaluation, diverse disease information, effective use of external knowledge, and tasks closer to real-world scenarios. The dataset is generated using the Llama-3.1-70B-Instruct model, where the questions describe the patients symptoms and background, and the answers provide the correct diagnosis based on the ground truth.