Using the Pinecone SDK with Asyncio support and FastAPI

This is a sample app to demonstrate how to use the Pinecone Python SDK with asyncio support. This is a simple Next.js app with a FastAPI backend.

Getting started

Prerequisites

• A Pinecone account. Create your free account here if you don't already have one.
• A Pinecone API key by following these steps.
• Python 3.12 or higher.
• Node.js 20 or higher.

Setup

Create Python virtual environment

Create and activate a Python virtual environment:

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

Install Python dependencies

pip3 install -r requirements.txt

Create indexes and load data

If you don't have both a dense index and a sparse index loaded with your own data, you can load test data using the load-data.py in the scripts directory. This will create two indexes:

1. a dense index using NVIDIA's llama-text-embed-v2, a state-of-the-art embedding model available natively in Pinecone Inference
2. a sparse index using Pinecone's own pinecone-sparse-english-v0 embedding model

Once the indexes are created, it will chunk and upsert data from the GEM/sportsett_basketball dataset on Huggingface, converting the text to vectors automatically using the hosted embedding model.

Setup the script environment

Set up your scripts environment variables by copying the .env.example file in the scripts directory to .env and replacing with your environment values.

Run the script

From the scripts directory, run:

python3 load-data.py

Setup the app environment

Set up your app environment variables by copying the .env.example file in the project root to .env and replacing with your environment values. You'll need to grab the index host URLs either from the Pinecone console or from the Pinecone API using describe_index as detailed here.

Install Node dependencies

From the project root, run:

npm install

Run the app

From the project root, run:

npm run dev

Head over to http://localhost:3000 to view the app.

The FastAPI server will be running on http://127.0.0.1:8000.