1. Create and active a Python virtual environment by running the following commands:

python3 -m venv ./.venv
source ./.venv/bin/activate

2. Install required dependencies by runing the following command:

pip install singlestoredb langchain langchain-community langchain-openai openai pypdf ijson "fastapi[standard]" python-dotenv

3. Create a .env file in the root of the project based on the .env.example file.
4. Place a PDF, Markdown, or TXT file that you want to use as your knowledge base into the directory you just created.
5. Create tables by running the following command:

python ./create_tables.py

6. Create embeddings by running the following command:

python ./create_embeddings.py

7. Load embeddings by running the following command:

python ./load_embeddings.py

8. Run the app:

python ./app.py

9. Open http://localhost:8000/docs in your browser and test the /ask API endpoint.

Start With SingleStore – $600 in Free Credits