Introduction
1. Introduction
Laurence: Hi, I'm Laurence, and I'm a developer advocate on Google Brain, focused on TensorFlow. In this course, you're gonna be learning the considerations behind architecting and implementing production machine learning systems. Now one key consideration of this, of course, is performance. In this module, you'll learn how to identify performance considerations for machine learning models. Now machine learning models are not all identical. For some models, you'll be focused on improving IO performance, and on others, you'll be focused on squeezing out more computational speed. Depending on what your focus is, you will need different ML infrastructure. Whether you decide to scale out with multiple machines, or scale up on a single machine with a GPU or TPU. Sometimes you might even need to do both by using a machine with multiple accelerators attached to it. Now it's not just a hardware choice. The hardware you select will also inform your choice of a distribution strategy.2. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.