README.md

metadata

title: Dupper Analytics QA
emoji: ☁️
colorFrom: purple
colorTo: pink
sdk: gradio
sdk_version: 3.39.0
app_file: app.py
pinned: true
license: bigscience-openrail-m

Overview

This is a basic Question-Answer style of chain using a small vector database created from the content in the Dupper Analytics website. (The website itself isn't very comprehensive, so don't expect the model to have immense knowledge) The question is embedded as a vector, the database retrieves the nearest vectors in the database, and that is then input into GPT-3.5-turbo to generate a cohesive answer. This entire workflow is visible via the app.py file.

Citation Information

The content used to generate the database

Other Notes

This demo was created to show how easily this model can be applied to various databases