Skip to content

A kedro-plugin for integration of mlflow capabilities inside kedro projects (especially machine learning model versioning and packaging)

License

Notifications You must be signed in to change notification settings

Galileo-Galilei/kedro-mlflow

Repository files navigation

General informations

Python Version License Code Style: Black SemVer


Package manager Software repository Latest release Total downloads
pip PyPI PyPI version Downloads
conda conda-forge conda version Downloads

Code health


Branch Tests Coverage Links Documentation Deployment Activity
master test codecov links Documentation publish commit

If you like the repo, please give it a ⭐

What is kedro-mlflow?

kedro-mlflow is a kedro-plugin for lightweight and portable integration of mlflow capabilities inside kedro projects. It enforces Kedro principles to make mlflow usage as production ready as possible. Its core functionalities are :

  • versioning: kedro-mlflow intends to enhance reproducibility for machine learning experimentation. With kedro-mlflow installed, you can effortlessly register your parameters or your datasets with minimal configuration in a kedro run. Later, you will be able to browse your runs in the mlflow UI, and retrieve the runs you want. This is directly linked to Mlflow Tracking.
  • model packaging: kedro-mlflow intends to be be an agnostic machine learning framework for people who want to write portable, production ready machine learning pipelines. It offers a convenient API to convert a Kedro pipeline to a model in the mlflow sense. Consequently, you can API-fy or serve your Kedro pipeline with one line of code, or share a model with without worrying of the preprocessing to be made for further use. This is directly linked to Mlflow Models.

How do I install kedro-mlflow?

Important: kedro-mlflow is only compatible with kedro>=0.16.0 and mlflow>=1.0.0. If you have a project created with an older version of Kedro, see this migration guide.

kedro-mlflow is available on PyPI, so you can install it with pip:

pip install kedro-mlflow

If you want to use the most up to date version of the package which is under development and not released yet, you can install the package from github:

pip install --upgrade git+https://github.com/Galileo-Galilei/kedro-mlflow.git

I strongly recommend to use conda (a package manager) to create an environment and to read kedro installation guide.

Getting started

The documentation contains:

  • A "hello world" example which demonstrates how you to setup your project, version parameters and datasets, and browse your runs in the UI.
  • A section for advanced machine learning versioning to show more advanced features (mlflow configuration through the plugin, package and serve a kedro Pipeline...)
  • A section to demonstrate how to use kedro-mlflow as a machine learning framework to deliver production ready pipelines and serve them. This section comes with an example repo you can clone and try out.

Some frequently asked questions on more advanced features:

Release and roadmap

The release history centralizes packages improvements across time. The main features coming in next releases are listed on github milestones. Feel free to upvote/downvote and discuss prioritization in associated issues.

Disclaimer

This package is still in active development. We use SemVer principles to version our releases. Until we reach 1.0.0 milestone, breaking changes will lead to <minor> version number increment, while releases which do not introduce breaking changes in the API will lead to <patch> version number increment.

The user must be aware that we will not reach 1.0.0 milestone before Kedro does (mlflow has already reached 1.0.0). That said, the API is considered as stable from 0.8.0 version and user can reliably consider that no consequent breaking change will happen unless necessary for Kedro compatibility (e.g. for minor or major Kedro version).

If you want to migrate from an older version of kedro-mlflow to most recent ones, see the migration guide.

Can I contribute?

We'd be happy to receive help to maintain and improve the package. Any PR will be considered (from typo in the docs to core features add-on) Please check the contributing guidelines.

Main contributors

The following people actively maintain, enhance and discuss design to make this package as good as possible:

Many thanks to Adrian Piotr Kruszewski for his past work on the repo.