TensorFlow.js was originally designed to achieve the high performance that's required for us to run machine learning applications while using hardware acceleration mechanisms such as WebGL. Having said that, there are several points to take note of so that we can make the most of these mechanisms and gain higher efficiency. One of the biggest advantages of TensorFlow.js over other frontend machine learning frameworks is the performance acceleration that's leveraged by several of its backend implementations. Thanks to these accelerators, applying TensorFlow.js can help us achieve competitive performance. We will learn how we can make use of these backend implementations to pursue high performance, as well as some useful tips we can use to tune the application we've written in TensorFlow.js.
In this chapter, we&apos...
