Chapter 8. Training Models

“Ask not for a lighter burden, but for broader shoulders.”

—Jewish proverb

While the supply of impressive models and data will continue to grow and overflow, it’s reasonable that you’ll want to do more than just consume TensorFlow.js models. You’ll come up with the idea that’s never been done before, and there won’t be an off-the-shelf option that day. It’s time for you to train your own model.

Yes, this is the task where the best minds in the world compete. While libraries could be written about the math, strategy, and methodology of training models, a core understanding will be vital. It’s crucial that you become familiar with the basic concepts and benefits of training a model with TensorFlow.js to take full advantage of the framework.

We will:

  • Train your first model in JavaScript code

  • Advance your understanding of model architecture

  • Review how to keep track of status during training

  • Cover some fundamental concepts of training

When you finish this chapter, you’ll be armed with a few ways of training a model and a better understanding of the process of using data to make a machine learning solution.

Training 101

It’s time to peel back the magic and train a model with JavaScript. While Teachable Machine is a great tool, it’s limited. To really empower machine learning, you’re going to have to identify the problem you want to solve and then teach a machine to find the patterns for a solution. To do this, we’ll view a problem through the ...

Get Learning TensorFlow.js now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.