/MLDeploy

How to deploy your Machine Learning models to production

Primary LanguageJupyter NotebookMIT LicenseMIT

MLDeploy

How to simply deploy your Machine Learning models to production

Getting Started

These instructions will get you a copy of the code up and running on your local host for development and testing purposes. Deployment will focus on one platform of my choice.

Prerequisites

Things you will need to bring the project up on your local machine

Docker (Not necessarily but advised)
Jupyter notebook
Python3
Text editor

Installing

A step by step series of getting a development env running on your local machine

Windows

mkdir Big4
cd Big4
virtualenv --no-site-packages venv
venv\Scripts\activate
git clone https://github.com/0x6f736f646f/MLDeploy.git
cd Big4Trendanalysis

Unix

mkdir Big4
cd Big4
virtualenv --no-site-packages venv
.venv/bin/activate
git clone https://github.com/0x6f736f646f/MLDeploy.git
cd Big4Trendanalysis

Installing requirements

pip install -r requirements

Running webapp

python3 web.py

Running webapi

python3 api.py

Dockerising your app

docker build -t mldeploy:1.0 .
  • -t is to tag the image being built
  • mldepoly is the image name your can replace it with your own image name
  • 1.0 is the version

Running docker web app container

docker run -d -p 4000:5000 big4trendanalysis:10.1 python3 web.py
  • -d is running it as a daemon
  • -p is for port mapping (We are mapping 5000 from docker container to 4000 to our localhost)
  • mldeploy this is the image name we built
  • 1.0 this is the version of the image we built
  • python3 web.py this is passing a command to the container

Running docker web api container

docker run -d -p 4000:5000 big4trendanalysis:10.1 python3 web.py
  • -d is running it as a daemon
  • -p is for port mapping (We are mapping 5000 from docker container to 4000 to our localhost)
  • mldeploy this is the image name we built
  • 1.0 this is the version of the image we built
  • python3 api.py this is passing a command to the container

Web app

Windows

Find your docker ip

docker-machine.exe ip

Then go to http:// docker-machine ip:4000

Otherwise

Go to http://localhost:4000 to find the web app or api

Deployment

Procedure on how to deploy on heroku as a live system. I chose installation through the command line because it was easier for me than the web based view.

install heroku cli for windows and linux users

heroku login
heroku create --region eu your_appname # creates app in eu region, common regions: eu, us
heroku buildpacks:set heroku/python # set python buildpack
git push heroku master # deploy app to heroku
heroku logs --tail # If for some reason it’s not working, check the logs

Built With

  • Docker - Container development tool
  • Flask - The web framework used
  • Heroku - Platform

Contributing

Please read CONTRIBUTING.md for details on our code of conduct, and the process for submitting pull requests to us.

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

  • AI saturday Kenya