NJUxlj/Chinese-MedQA-Qwen2
基于Qwen2+SFT+DPO的医疗问答系统，项目中使用了自定义的 SFTTrainer/DPOTrainer/TRPOTrainer用于训练，其次，项目还调用各种知识库工具（neo4j, milvus, LDA, 等）进行自动化训练数据生成。另外，使用 vllm 用于推理和部署训好的模型, 该模型会通过 vllm API 来接入一个基于 embedder + Reranker 的 RAG 系统。另外还参考 MDAgents 论文实现了一个多智能体会诊系统，同样也支持 vllm api 接入。
This project helps medical professionals or healthcare organizations create specialized AI assistants for medical inquiries. It takes existing medical knowledge, patient dialogue data, and expert feedback to fine-tune a powerful language model. The result is a Western medicine-focused AI that can diagnose diseases and provide informed responses, accessible through an intuitive question-and-answer system.
Use this if you need to build and deploy a highly accurate, specialized AI model to answer complex medical questions based on Western medicine knowledge.
Not ideal if you are looking for a general-purpose chatbot or an AI system that incorporates traditional Chinese medicine or other alternative medical practices.
Stars
64
Forks
9
Language
Python
License
—
Category
Last pushed
Jan 04, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/NJUxlj/Chinese-MedQA-Qwen2"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
ImprintLab/Medical-Graph-RAG
A Graph RAG System for Evidenced-based Medical Information Retrieval [ACL 2025]
Warma10032/cyber-doctor
赛博医生项目——”赛博华佗“，基于多模态大模型的多功能智能体，一键搭建本地多模态大模型。接入医疗健康相关的知识图谱和知识库后可以进行疾病初诊，病历分析，专业知识问答等功能，成为你的私人医生。赛博...
Tyler-KD/multi-model-AI-assistant-medical-bot
RAG-enabled multi-agentic system for medical diagnosis and assistance
slinusc/medical_RAG_system
Enhancing Medical Question-Answering System through Advanced Information Retrieval Strategies
avnlp/biothink
Self-Reflective Question Answering for Biomedical Reasoning