Oreilly – Distributed Machine Learning Patterns, Video Edition 2024-1

Distributed Machine Learning Patterns Video Edition. This course explores practical patterns for scaling machine learning from your personal laptop to a distributed cluster. In this course, participants will learn techniques and expert tips for tackling the challenges of scaling machine learning systems.

Distributed machine learning systems allow developers to manage very large datasets across multiple clusters, leverage automation tools, and take advantage of hardware accelerations. Distributed Machine Learning Patterns reveals best practice techniques and insider tips for tackling the scalability challenges of machine learning systems. In this book, Yuan Tang, project lead for Argo and Kubeflow, shares patterns, examples, and insights gained from his experience migrating an ML model from a single machine to a distributed cluster.

What you will learn

• You will apply distributed systems patterns to build scalable and reliable machine learning projects. You will build ML pipelines with data ingestion, distributed training, model serving, and more.
• You will automate ML tasks with Kubernetes, TensorFlow, Kubeflow, and Argo Workflows.
• You will make the right choices between different patterns and approaches.
• You will manage and monitor machine learning workloads at scale.

This course is suitable for people who:

• There are data analysts and engineers who are familiar with the basics of machine learning, Bash, Python, and Docker.

Distributed Machine Learning Patterns Video Edition Course Details

• Publisher: Oreilly
• Instructor: Yuan Tang
• Training level: Beginner to advanced
• Training duration: 6 hours and 21 minutes

Course topics

• Part 1. Basic concepts and background
• Chapter 1. Introduction to distributed machine learning systems
• Chapter 1. Distributed systems
• Chapter 1. Distributed machine learning systems
• Chapter 1. What we will learn in this book
• Chapter 1. Summary
• Part 2. Patterns of distributed machine learning systems
• Chapter 2. Data ingestion patterns
• Chapter 2. The Fashion-MNIST dataset
• Chapter 2. Batching pattern
• Chapter 2. Sharding pattern: Splitting extremely large datasets among multiple machines
• Chapter 2. Caching pattern
• Chapter 2. Answers to exercises
• Chapter 2. Summary
• Chapter 3. Distributed training patterns
• Chapter 3. Parameter server pattern: Tagging entities in 8 million YouTube videos
• Chapter 3. Collective communication pattern
• Chapter 3. Elasticity and fault-tolerance pattern
• Chapter 3. Answers to exercises
• Chapter 3. Summary
• Chapter 4. Model serving patterns
• Chapter 4. Replicated services pattern: Handling the growing number of serving requests
• Chapter 4. Sharded services pattern
• Chapter 4. The event-driven processing pattern
• Chapter 4. Answers to exercises
• Chapter 4. Summary
• Chapter 5. Workflow patterns
• Chapter 5. Fan-in and fan-out patterns: Composing complex machine learning workflows
• Chapter 5. Synchronous and asynchronous patterns: Accelerating workflows with concurrency
• Chapter 5. Step memoization pattern: Skipping redundant workloads via memoized steps
• Chapter 5. Answers to exercises
• Chapter 5. Summary
• Chapter 6. Operation patterns
• Chapter 6. Scheduling patterns: Assigning resources effectively in a shared cluster
• Chapter 6. Metadata pattern: Handle failures appropriately to minimize the negative effect on users
• Chapter 6. Answers to exercises
• Chapter 6. Summary
• Part 3. Building a distributed machine learning workflow
• Chapter 7. Project overview and system architecture
• Chapter 7. Data ingestion
• Chapter 7. Model training
• Chapter 7. Model serving
• Chapter 7. End-to-end workflow
• Chapter 7. Answers to exercises
• Chapter 7. Summary
• Chapter 8. Overview of relevant technologies
• Chapter 8. Kubernetes: The distributed container orchestration system
• Chapter 8. Kubeflow: Machine learning workloads on Kubernetes
• Chapter 8. Argo Workflows: Container-native workflow engine
• Chapter 8. Answers to exercises
• Chapter 8. Summary
• Chapter 9. A complete implementation
• Chapter 9. Model training
• Chapter 9. Model serving
• Chapter 9. The end-to-end workflow
• Chapter 9. Summary

Course images

Sample course video

Download link

Download file – 868 MB

File(s) password: www.downloadly.ir

File size

868 MB