- pip3 install -r requirements.txt
- python app.py
To scrape the data, run the scraper.py
script and save it in the following CSV format:
title, url, content
As GPT-3.5 only supports 4k tokens, we need to split the data. To do this, run the splitter.py
script.
Run the insert_emebedding.py
script to create embeddings for the data and insert to table in supabase
For quick and easy to get started, Create a database in supabase and run the schema.sql
script in Supabase. Enable vector extension
Set the following environment variables in your ~/.bash_profile
or ~/.bashrc
file:
SUPABASE_FN_URL
(tim_search function URL)SUPABASE_URL
SUPABASE_KEY
OPENAI_API_KEY
Thanks to @tferriss for all the podcast
It was his podcast I first ever listen almost
If you have any questions, feel free to reach out to me on Twitter @_rshiva
I kept it as simple as possible for users to jump in OPENAI.