Deploying Machine Learning Models using FastAPI and Docker

This example focuses on the following:

• Building a machine learning model with Scikit-learn (model train and save)
• Creating a FastAPI to serve predictions
• Containerizing the API using Docker

Installation

• Create a virtual environment

python3 -m venv ~/opt/MLFlow

• Activate it and install necessary packages using requirements.txt

source ~/opt/MLFlow/bin/activate
pip install -r requirements.txt

• Build the Docker image by running the following docker build command:

docker build -t house-price-prediction-api .

• Next run the Docker container:

docker run -d -p 80:80 house-price-prediction-api

Your API should now be running and accessible at http://127.0.0.1:80.

You can use curl or Postman to test the /predict endpoint by sending a POST request. Here’s an example request:

curl -X 'POST' \
  'http://127.0.0.1:80/predict' \
  -H 'Content-Type: application/json' \
  -d '{
  "MedInc": 3.5,
  "AveRooms": 5.0,
  "AveOccup": 2.0
}'

Also, you can run the following command to execute the docker inside

docker exec -it [docker-id] /bin/bash

To find the Docker ID:

docker ps