Tutorials and Examples#

The following tutorials provide a hands-on introduction to using MLRun to implement a data science workflow and automate machine-learning operations (MLOps).

Make sure you start with the Quick start tutorial to understand the basics

Introduction to MLRun - Use serverless functions to train and deploy models

Targeted Tutorials#

Each of the following tutorials is a dedicated Jupyter notebook. You can download them by clicking the download icon at the top of each page.

Train, compare, and register Models

Demo of training ML models, hyper-parameters, track and compare experiments, register and use the models.

Serving pre-trained ML/DL models

How to deploy real-time serving pipelines with MLRun Serving and different types of pre-trained ML/DL models.

Projects & automated ML pipeline

How to work with projects, source control (git), CI/CD, to easily build and deploy multi-stage ML pipelines.

Real-time monitoring & drift detection

Demonstrate MLRun Serving pipelines, MLRun model monitoring, and automated drift detection.

Add MLOps to existing code

Turn a Kaggle research notebook to a production ML micro-service with minimal code changes using MLRun.

Basic feature store example (stocks)

Understand MLRun feature store with a simple example: build, transform, and serve features in batch and in real-time.

Batch inference and drift detection

Use MLRun batch inference function (from MLRun Function Hub), run it as a batch job, and generate drift reports.

Advanced real-time pipeline

Demonstrates a multi-step online pipeline with data prep, ensemble, model serving, and post processing.

Feature store end-to-end demo

Use the feature store with data ingestion, model training, model serving, and automated pipeline.

End to End Demos#

You can find the different end-to-end demos in the MLRun demos repository: github.com/mlrun/demos.

Running the demos in Open Source MLRun#

By default, these demos work with the online feature store, which is currently not part of the Open Source MLRun default deployment:

  • fraud-prevention-feature-store

  • network-operations

  • azureml_demo