/doctor-llm

Primary LanguagePythonMIT LicenseMIT

Doctor LLM

Description

This is small LLM project using Langchain with RAG and ReAct acting as a Doctor. The RAG use MASHQA dataset as the reference for the model to be better at answering questions.

Issues

  • Model still sometime doesn't follow the rules set by ReAct prompt
  • When using with ReAct, RAG sometimes only return question from the vectordb instead of both question and answer. Which result into the final answer being a question instead of an answer toward the user.
  • ReAct doesn't work well with smaller model.

Requirements

  • langchain
  • langchain-community
  • langchainhub
  • langchain-openai
  • langchain-huggingface
  • bs4
  • streamlit
  • faiss-cpu

Example

Q&A-1 Q&A-2 Q&A-3

Running

pip install -r requirements.txt

streamlit run app.py 

References