MLRun Package Documentation

Introduction

MLRun is a generic and convenient mechanism for data scientists and software developers to describe and run tasks related to machine learning (ML) in various, scalable runtime environments and ML pipelines while automatically tracking executed code, metadata, inputs, and outputs. MLRun integrates with the Nuclio serverless project and with Kubeflow Pipelines.

MLRun features a Python package (mlrun), a command-line interface (mlrun), and a graphical user interface (the MLRun dashboard).