From Jupyter to Production

Production-ready Data Science Projects

This repository contains material for the workshop "From Jupyter to Production". The goal of the workshop is to get a glimpse of production-readiness for data science and machine learning projects.

With the introductory Jupyter notebooks and the exercises found in the notebooks directory, you will learn how to

• Versioning your data and models with DVC
• Build pipelines with Dagster
• Track experiments with MLflow
• Deploy your model with FastAPI

Having installed docker, you can use JupyterLab for the exercises.

Start JupyterLab

First clone the repository

git clone https://github.com/codecentric/from-jupyter-to-production-workshop
cd from-jupyter-to-production-workshop

and then execute the command

docker compose up -d

You can now use JupyterLab in your browser: http://localhost:8888

Docker Images

If you want to pull the docker images separately

docker pull codecentric/from-jupyter-to-production-baseimage

You will find the source for the docker images here:

http://github.com/codecentric/from-jupyter-to-production-baseimage

Data Sources

https://archive.ics.uci.edu/ml/datasets/wine+quality