Machine Learning in Production: MLflow and Model Deployment

Machine Learning in Production: MLflow and Model Deployment

Summary

In this 1-day course, data scientists and data engineers learn best practices for managing experiments, projects, and models using MLflow. Students build a pipeline to log and deploy machine learning models, as well as explore common production issues faced when deploying machine learning solutions and monitoring these models once they have been deployed into production.

Description

 

This course is separated into two main components. The first uses MLflow as the backbone for machine learning development and production. This includes tracking the machine learning lifecycle, packaging projects for deployment, using the MLflow model registry, and more. The second component looks at various production issues, the four main deployment paradigms, monitoring, and alerting. Depending on the desires of the class, numerous electives are also available on the various MLflow functionality and deployment scenarios.

 

By the end of this course, you will have built the infrastructure to manage the development, deployment, and monitoring of models in production. This course is taught entirely in Python.

Duration

8 hours

Objectives

Upon completion, students should be able to:

  • Track machine learning experiments to organize the machine learning life cycle
  • Create, organize, and package machine learning projects with a focus on reproducibility and collaborating with a team
  • Develop a generalizable way of handling machine learning models created in and deployed to a variety of environments
  • Explore the various production issues encountered in deploying and monitoring machine learning models
  • Introduce various strategies for deploying models using batch, streaming, and real-time
  • Explore solutions to drift and implement a basic retraining method and two ways of dynamically monitoring drift

Audience

  • Data Scientist

  • Machine Learning Engineer

  • Data Engineer

Prerequisites

  • Intermediate experience using Python/pandas
  • Working knowledge of machine learning and data science (scikit-learn, TensorFlow, etc.)
  • Familiarity with Apache Spark
  • Basic familiarity with object storage, databases, and networking

 

Outline

 

Day #1 AM

Duration

Modules

Description

30 min

Introductions & Setup

Registration, Courseware & Q&As

40 min

Experiment Tracking

Tracking experiments to organize the machine learning life cycle

10 min

Break

 

30 min

Packaging ML Projects

Creating, organizing, and packaging machine learning projects

30 min

Multistep Workflows

Managing the complexity of multistep machine learning projects using multistep workflows

10 min

Break

 

30 min

Model Management

A generalizable way of handling machine learning models created in and deployed to a variety of environments

180 min

   

Day #1 PM

Duration

Modules

Description

30 min

Model Registry

How to manage models using the MLflow model registry

40 min

Production Issues

Various production issues encountered in deploying and monitoring machine learning models

15 min

Break

 

30 min

Batch Deployment

Various strategies for deploying models using batch including pure Python, Spark, and on the JVM

30 min

Streaming Deployment

How to perform inference on a stream of incoming data

10 min

Break

 

40 min

Real Time Deployment

Real time deployment with a focus on RESTful services

10 min

Break

 

30 min

Drift Monitoring

Explore solutions to concept and data drift

30 min

Alerting

Alerting strategies using email and REST integration

Upcoming Classes

Date
Time
Location
Price
Jan 25 - 26
9:00 AM - 1:00 PM
Pacific Standard Time
Online - Virtual - Americas (half-day schedule)
$ 1000.00 USD
Apr 26 - 27
9:00 AM - 1:00 PM
Pacific Daylight Time
Online - Virtual - Americas (half-day schedule)
$ 1000.00 USD
Jun 7 - 8
9:00 AM - 1:00 PM
Pacific Daylight Time
Online - Virtual - Americas (half-day schedule)
$ 1000.00 USD
Jul 12 - 13
9:00 AM - 1:00 PM
Pacific Daylight Time
Online - Virtual - Americas (half-day schedule)
$ 1000.00 USD