Fast-ML-API

• ML model inference serving template using Fast API and Docker

Other Features to consider

• Using ujson or rjson.
• Token Auth.
• Exception Handling.
• Model Versioning with APIRoute.
• Message Broker to support Request Batching.

Instructions

1. Clone the repository OR Use Docker

manually

cd fast-ml-api
bash run.sh
python tests/api_test.py
uvicorn project_api.main:app

(OR)

build the docker-image & run the container

bash deploy.sh

2. Go to docs to get available apis

http://127.0.0.1:8000/docs

3. Stop the server

ctrl + c