ChatMed_Consult_Dataset

ChatMed-Dataset仓库专注于为中文医学大型语言模型注入医学知识。它包含11万多条由OpenAI的GPT-3.5引擎生成的中文医疗问答对，问题来源于互联网医疗咨询，反映了真实用户的医疗需求。此数据集专为微调预训练语言模型而设计，以期在自动医疗咨询方面表现更佳。数据以JSON行格式存储，方便使用，并采用CC-BY 4.0授权许可。请用户谨慎使用，并提出新的方法来过滤或改进其中的不完善之处。

The ChatMed-Dataset repository focuses on infusing medical knowledge into Chinese medical large language models. It contains over 110,000 Chinese medical Q&A pairs generated by OpenAI's GPT-3.5 engine, with questions sourced from online medical consultations that reflect the real medical needs of users. This dataset is specifically designed for fine-tuning pre-trained language models to achieve better performance in automated medical consultation tasks. The data is stored in JSON Lines format for ease of use, and is licensed under CC-BY 4.0. Users are requested to use it with caution and propose novel methods to filter or improve its existing imperfections.

搜集汇总

背景与挑战

背景概述

ChatMed_Consult_Dataset是一个包含11万条中文医疗问答对的数据集，由GPT-3.5生成，旨在为中文医学大型语言模型提供知识支持。数据来源于真实医疗咨询，以JSON行格式存储，适用于模型微调，但需注意潜在的错误和偏见。