Book Image

Distributed Machine Learning with Python

By : Guanhua Wang
Book Image

Distributed Machine Learning with Python

By: Guanhua Wang

Overview of this book

Reducing time cost in machine learning leads to a shorter waiting time for model training and a faster model updating cycle. Distributed machine learning enables machine learning practitioners to shorten model training and inference time by orders of magnitude. With the help of this practical guide, you'll be able to put your Python development knowledge to work to get up and running with the implementation of distributed machine learning, including multi-node machine learning systems, in no time. You'll begin by exploring how distributed systems work in the machine learning area and how distributed machine learning is applied to state-of-the-art deep learning models. As you advance, you'll see how to use distributed systems to enhance machine learning model training and serving speed. You'll also get to grips with applying data parallel and model parallel approaches before optimizing the in-parallel model training and serving pipeline in local clusters or cloud environments. By the end of this book, you'll have gained the knowledge and skills needed to build and deploy an efficient data processing pipeline for machine learning model training and inference in a distributed manner.
Table of Contents (17 chapters)
1
Section 1 – Data Parallelism
6
Section 2 – Model Parallelism
11
Section 3 – Advanced Parallelism Paradigms

Collective communication

Besides the popular All-Reduce function, collective communication has a wide family of message-passing functions. We will discuss several of the most important collective communication functions in this section, namely Broadcast, Gather, and All-Gather.

Broadcast

Broadcast is also widely used in the All-Reduce architecture. For example, we can use Broadcast to distribute the initial model weights among all the workers. The following diagram shows an example of the Broadcast function in a three-workers setting:

Figure 2.13 – Broadcast in a three-workers setting

As we can see, initially, Worker 1 holds a value of a. Worker 2 holds a value of b, while Worker 3 holds a value of c. Here, we conduct a broadcast from Worker 1 to all the other workers.

After this broadcast operation, Worker 1 still holds a value of a. However, Worker 2 now holds values a and b. On the other hand, Worker 3 holds values a and c.

For this...