Teach on Udemy

Turn what you know into an opportunity and reach millions around the world.

Learn More

Your cart is empty.

Keep shopping

MLflow in Action - Master the art of MLOps using MLflow tool

Name: MLflow in Action - Master the art of MLOps using MLflow tool
Rating: 4.6 (1138 reviews)

A master guide to unleash the full potential of MLflow to optimize MLOps. Streamline MLOps workflows using MLflow tool

Created byJ Garg - Real Time Learning

Last updated 6/2025

English

What you'll learn

Explore the fundamentals of MLOps and how it overcomes the challenges inherent in the traditional ML lifecycle.
Gain a deep understanding of MLflow and the role of its 4 components in managing the end-to-end Machine learning operations (MLOps).
Learn how to efficiently Track experiments, Package code, Register and reproduce models in the realm of MLOps using MLflow tool.
A range of MLflow logging functions to effectively track and record experiments, runs, artifacts, parameters, code, metrics etc.
MLflow Tracking - To log, organize, and compare Machine learning experiments effortlessly.
MLflow Model - For efficient model packaging into distinct flavors allowing to streamline model deployment and integration into production systems.
MLflow Project - To create structured, reproducible, and easily shareable Machine Learning workflows.
MLflow Registry - For efficient model management, version tracking in order to maintain model quality and performance over time.
A complete end-to-end ML project demonstrating MLflow integration with AWS cloud.
Build, Train, Test and Deploy a Machine learning model in AWS cloud using AWS Sagemaker and MLflow.

Course content

16 sections • 87 lectures • 9h 15m total length

What is MLOps4:09
Definition of MLOps
Traditional Machine Learning Lifecycle4:50
What is the traditional machine learning lifecycle followed in most of the Machine learning projects.
Challenges in traditional ML lifecycle - Part 16:22
What are the problems in existing Machine learning projects that led to 80% of non-productionization of models. Part 1
Challenges in traditional ML lifecycle - Part 28:39
What are the problems in existing Machine learning projects that led to 80% of non-productionization of models. Part 2
How MLOps address the challenges5:05
How MLOps address the traditional Machine learning lifecycle challenges.

Setting and Getting Tracking Uri - Part 14:28
Setting and Getting Tracking Uri - Part 23:07
Implement the set_tracking_uri and get_tracking_uri functions of MLflow tracking component
Create MLflow experiment function9:15
How to create an experiment in MLflow
Set MLflow experiment function3:17
How to implement set_experiment() function in MLflow (an MLOps tool)
Start and End run functions - Part 15:09
How to implement Start and End run functions using MLflow MLOps tool.
Start and End run functions - Part 24:53
How to implement Start and End run functions using MLflow MLOps tool.
Active & Last active run functions4:23
How to implement Active run & Last active run functions in MLflow
Log multiple parameters & metrics function7:42
Log multiple artifacts function9:10
How to Log multiple artifacts function in MLflow
Set tags function7:16
How to set a single tag and multiple tags at once using set_tag and set_tags MLflow functions
Quiz 1
Assignment 1

What is MLflow Tracking server5:36
Introduction to MLflow Tracking server. This lesson includes the fundamentals around MLflow Tracking server.
Implement Tracking Server in MLflow5:42
How to add a Tracking server in MLflow program.
Local Tracking server scenarios6:21
This video describes some of the Local Tracking server scenarios in MLflow.
Remote Tracking server scenarios7:54
This video describes some of the Remote Tracking server scenarios in MLflow

Introduction to Model component5:23
Introduction to MLflow Model component
Model components6:41
Detailed explanation to some of the important terminologies in MLflow model component.
Storage format7:47
MLflow packages the model in different formats known as MLflow flavors
MLmodel file explained7:56
In this video, MLmodel file component is explained for sklearn library.
What are Model Signatures5:54
A quick knowledge on Model Signatures and Input examples in Machine learning and what type of model signatures are supported by MLFlow.
What is Signature Enforcement in MLflow4:46
Concept of Signature enforcement explained in MLflow
Log Signatures & Input examples in MLflow11:09
Log model signatures and input examples in a MLflow program
Model API (save_model, log_model) explained10:27
Quiz 2

What is Model customization in MLflow3:21
Introduction to Model customization in MLflow
Implement Custom Python model - Part 14:28
How to implement a Custom Python model using Python flavor in MLflow - Part 1
Implement Custom Python model - Part 26:08
How to implement a Custom Python model using Python flavor in MLflow - Part 2
Implement Custom Python model - Part 34:55
How to implement a Custom Python model using Python flavor in MLflow - Part 3
Loading the customized Python model3:06
Loading the customized Python model in MLflow
Custom Flavors in MLflow8:28
What are Custom Flavors in MLflow and how to implement them.

Introduction to Model evaluation in MLflow3:20
Introduction to Model evaluation in MLflow
Evaluate function parameters9:31
What are the parameters of mlflow.evaluate function in MLflow
Implement evaluate() function6:26
Implementation of mlflow.evaluate method in the ongoing mlflow program.
Comparing Runs of an Experiment7:37
Explanation of all the comparison plots while comparing the runs in an mlflow experiment.
Create Custom metrics and artifacts7:00
Setting Validation thresholds in evaluate()7:49
Assignment 2

Requirements

Basic Machine Leaning knowledge
Basic Python knowledge

Description

Why MLOps ?

MLOps is the backbone of modern Machine learning workflows. It solves the pressing problem of operationalizing the ML models in production systems. Pushing the ML models to production which could traditionally take months can now be operationalized in few days using MLOps tools.

As per the tech talks in market, 2024 is the year of MLOps and would become the mandate skill for Enterprise ML projects.

Why MLflow tool for MLOps ?

MLflow is the ultimate tool for MLOps because it streamlines the entire Machine learning lifecycle. It allows you to efficiently track experiments, package code, register versions and deploy models, all within one unified platform. Unlike other tools, MLflow simplifies the process, enabling you to transition from development to deployment seamlessly.

MLflow's popularity is evident from the thousands of organizations, ranging from startups to Fortune 500 companies, that have integrated MLflow into their MLOps workflows.

_____________________________________________________________________________________________________

What's included in this MLflow course ?

Understand MLOps basics, limitations of traditional ML lifecycles, how MLOps overcomes those limitations.
Complete MLflow concepts explained from Scratch to Real-Time implementation.
Learn in practical the 4 core components of MLflow - Tracking, Model, Project, and Registry.
Various logging functions in MLflow for precise tracking and recording of experiments, runs, artifacts, parameters, code, metrics, and more.
Learn to handle customized models using Python in MLflow.
Learn to interact with MLflow using MLflow library, UI, MLflow Client and CLI commands.
Learn Best practices and Optimization techniques to follow in Real-Time MLOps/MLflow Projects.

______________________________________________________________________________________________________

**Exclusive** - A complete end-to-end ML project demonstrating MLflow's integration with AWS cloud. Build, Train, Test, Deploy a Machine learning model in AWS cloud using AWS Sagemaker, Codecommit, Ec2, ECR, AWS S3, IAM etc services while leveraging MLflow tracking capabilities.

After completing this course, you can start working on any MLOps/MLflow project with full confidence.

Add-Ons

- Questions and Queries will be answered very quickly.

- Codes and references used in lectures are attached in the course for your convenience.

Who this course is for:

Data Scientists
Machine Learning Engineers
MLOps Engineers
Operations Engineers

MLflow in Action - Master the art of MLOps using MLflow tool

What you'll learn

Explore related topics

Course content

Introduction to MLOps5 lectures • 29min

Introduction to MLflow3 lectures • 22min

MLflow Tracking component4 lectures • 34min

MLflow Logging functions10 lectures • 59min

Launch multiple Experiments and Runs2 lectures • 18min

Autologging in MLflow3 lectures • 19min

Tracking Server of MLflow4 lectures • 26min

MLflow Model component8 lectures • 1hr

Handling Customized models in MLflow6 lectures • 30min

MLflow Model Evaluation6 lectures • 42min

Requirements

Description

Who this course is for: