DEPRECATED

This repository is no longer maintained and has been archived. Feel free to browse the code, but please migrate to other solutions.

python-flask-docker-sklearn-template

A simple example of python api for real time machine learning. On init, a simple linear regression model is created and saved on machine. On request arrival for prediction, the simple model is loaded and returning prediction.
For more information read this post

requirements

docker installed

Run on docker - local

docker build . -t {some tag name} -f ./Dockerfile_local
detached : docker run -p 3000:5000 -d {some tag name}
interactive (recommended for debug): docker run -p 3000:5000 -it {some tag name}

Run on docker - production

Using uWSGI and nginx for production
docker build . -t {some tag name}
detached : docker run -p 3000:80 -d {some tag name}
interactive (recommended for debug): docker run -p 3000:80 -it {some tag name}

Run on local computer

python -m venv env
source env/bin/activate
python -m pip install -r ./requirements.txt
python main.py

Use sample api

127.0.0.1:3000/isAlive
127.0.0.1:3000/prediction/api/v1.0/some_prediction?f1=4&f2=4&f3=4