MLflow

MLflow is an open-source platform for managing workflows and artifacts across the machine learning lifecycle, including experimentation, reproducibility, deployment, and a central model registry. ML has built-in integrations with a number of popular machine learning libraries but is library-agnostic and can be used with any library, algorithm, or deployment tool. MLflow provides four primary functions:

• Tracking experiments, recording and comparing parameters and results (MLflow Tracking)
• Packaging machine learning code in a reusable and reproducible form to share with other developers or transfer to production (MLflow Projects)
• Managing and deploying models from a variety of machine learning libraries to a variety of model serving and inference platforms (MLflow Models)
• Acting as a central store so developers can collaboratively manage the lifecycle of a machine learning workflow, including model versioning, stage transitions, and annotations (MLflow Model Registry)

MLflow works with any programming language, as all functions are accessible through a REST API and command line interface (CLI). MLflow also provides dedicated APIs for Python, R, and Java, for convenience. The first version of MLflow (0.1.0) was released on June 5, 2018.