Last updated: 2023-12-08
What is MLflow?
MLflow is an open-source platform to manage the lifecycle of ML models end to end. It tracks the code, data and results for each ML experiment, which means you have a history of all experiments at any time. A dream for every Data Scientist. Furthermore, MLflow is library-agnostic, which means you can use your favourite ML libraries like TensorFlow, PyTorch or scikit-learn.
As a Data Scientist, you spend a lot of time optimising ML models. The best models often depend on an optimal hyperparameter or feature selection, and it is challenging to find an optimal combination. Also, you have to remember all experiments, which is very time-consuming. MLflow is an efficient platform to address these challenges. The use of MLflow is significantly simplified by our MLflow Workspace.
The workspace consists of four services: a JupyterLab server, an MLflow tracking server, an artifact store and a backend store. All services run in a Docker container. You can develop your ML models in a Jupyter notebook on the JupyterLab server and track your experiments via MLflow. Moreover, you can view all experiments via the MLflow Tracking Server UI. For detailed information on the MLflow Workspace, we recommend reading our Towards Data Science article "Deploy Your Own MLflow Workspace On-Premise with Docker".
- Price: It's free! Just enter 0 for the price. You can enter a price of your choice if you want to support us.
- Environment: Docker
- Support for Apple Silicon, arm64 and amd64 architecture
- Services: JupyterLab server, MLflow tracking server with UI, artifact store (SFTP server) and backend store (PostgreSQL)
- Example Jupyter notebooks
Now it's time for a short demo.
You get a running MLflow environment with a tracking server UI and a JupyterLab server. How do you get it for free? Just put a 0 into the field "Name a fair price" and hit "I want this!".