Syllabus¶

This hands-on tutorial walks through the full MLOps lifecycle: from a local Jupyter notebook experiment to a reproducible pipeline, then to automated collaboration in the cloud, and finally to a deployed, monitored, and retrainable system.

The syllabus is organized into an introduction, five hands-on parts, and a conclusion:

• Introduction - Learn about the concept behind MLOps, the philosophy, and the tools used throughout, then set up your machine. A companion presentation gives an overview of the guide. A cheatsheet of useful terminal commands is available throughout the guide.
• Part 1 - Local training and evaluation - Learn how to train a model locally and evaluate it using DVC.
  • Chapter 1.1 - Run a simple ML experiment with Jupyter Notebook
  • Chapter 1.2 - Adapt and move the Jupyter Notebook to Python scripts
  • Chapter 1.3 - Initialize Git and DVC for local training
  • Chapter 1.4 - Reproduce the ML experiment with DVC
  • Chapter 1.5 - Track model evolution with DVC
• Part 2 - Move to the cloud - Learn how to collaborate online using Git, a CI/CD pipeline and CML.
  • Chapter 2.1 - Move the ML experiment code to the cloud
  • Chapter 2.2 - Move the ML experiment data to the cloud
  • Chapter 2.3 - Reproduce the ML experiment in a CI/CD pipeline
  • Chapter 2.4 - Track model evolution in the CI/CD pipeline with CML
  • Chapter 2.5 - Work efficiently and collaboratively with Git
• Part 3 - Serve and deploy - Learn how to serve and deploy the model using BentoML and Docker.
  • Chapter 3.1 - Save and load the model with BentoML
  • Chapter 3.2 - Serve the model locally with BentoML
  • Chapter 3.3 - Build and publish the model with BentoML and Docker locally
  • Chapter 3.4 - Build and publish the model with BentoML and Docker in the CI/CD pipeline
  • Chapter 3.5 - Deploy and access the model on Kubernetes
  • Chapter 3.6 - Continuous deployment of the model with the CI/CD pipeline
  • Chapter 3.7 - Use a self-hosted runner for the CI/CD pipeline
  • Chapter 3.8 - Train the model on a Kubernetes pod
• Part 4 - Monitor and maintain - Learn how to keep a model healthy in production using Evidently AI.
  • Chapter 4.1 - Log predictions and features locally
  • Chapter 4.2 - Detect drift locally with Evidently AI
  • Chapter 4.3 - Deploy and access the monitoring on Kubernetes
  • Chapter 4.4 - Trigger drift alerts with the CI/CD workflow
  • Chapter 4.5 - Review drift alerts and decide on action
• Part 5 - Label data and retrain - Learn how to label new data and retrain the model using Label Studio.
  • Chapter 5.1 - Set up Label Studio
  • Chapter 5.2 - Label new data with Label Studio
  • Chapter 5.3 - Link the model to Label Studio
  • Chapter 5.4 - Retrain the model from new data with DVC
• Conclusion - Clean up the resources you created along the way, then read the conclusion for a summary of what you have done.