This library records my experimentation with LLMs, the OpenAI API, and examining how they perform in police-like risk assessments. This outputs a Streamlit web app that is available here.
NBdev generated documentation is available through GitHub pages.
pip install police_risk_open_aiAmend the example.env file included in this repo, using an API key
from OpenAI, and save this as a new file, .env
You can use
query_llm
to query the API.
query_llm('Where does it rain in Spain?')'\n\nRainfall in Spain is'