simple-async-openai-assistant

Asynchronous FastAPI wrapper for AsyncOpenAI and OpenAI assistantAPI Resources

Usage

Local

We use Python3@11 for this project.

1. Configure your venv and install the requirements

python -m venv venv
source venv/bin/activate  # On Windows use `venv\Scripts\activate`
pip install -r requirements.txt

2. Add .secrets.yaml file to config/

---
OPENAI_API_KEY: <YOUR_OPENAI_API_KEY>

3. Start the ASGI uvicorn server:

uvicorn app.main:app --reload --host 0.0.0.0 --port 50050

4. Execute api calls concurrently

python async_query_requests.py <N> http://localhost:50050/api/query

Docker

1. Build the image:

docker build -t async-openai-assistant:latest .

2. Run the container:

docker run -p 50000:50050 async-openai-assistant:latest

3. Don't forget to set your config/.secrets.yaml to enable openai api

docker exec <container_id_or_name> sh -c 'echo "OPENAI_API_KEY: <YOUR_API_KEY_HERE>" > /app/config/.secrets.yaml'

4. Execute api calls concurrently

python async_query_requests.py <N> http://<YOUR_HOST>:50000/api/query