From Jupyter to Production

JupyterLab and Mlflow image for workshop: From Jupyter to Production - production-ready data science projects.

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

Build & Push

A github action is defined to push a new version of the image to Docker Hub every time a new git tag is pushed to the repository. The git tag is also used for the Docker images. An image tag with the git tag and the latest tag is published for each Docker build. No need to do anything locally, besides testing the build with docker build .

The build is quite memory heavy, so assign a good amount of memory towards the docker engine (minimum 4gb, better 6gb)

Run on Mac & Linux

Run in from-jupyter-to-production-workshop directory, containing the notebooks.

docker run -p 8888:8888 -v $(pwd)/notebooks:/workshop/notebooks radtkem/from-jupyter-to-production-jupyter

Run on Windows

Run in from-jupyter-to-production-workshop directory, containing the notebooks.

docker run -p 8888:8888 -v %cd%/notebooks:/workshop/notebooks radtkem/from-jupyter-to-production-jupyter