Chapter 7: Deployment and Model Maintenance

So far in this book, you have trained a wide variety of fastai models, including models trained with tabular datasets, models trained with text datasets, recommender systems, and models trained with image data. All the models that you have trained have been exercised in the context of Jupyter notebooks. Jupyter notebooks are great for training models and exercising them with a couple of test examples, but what about actually making your model useful? How do you make your model available to other people or applications to actually solve problems?

The process of making your deep learning models available to other people or applications is called deployment. In this chapter, we will go through recipes that show how to deploy your fastai models. The industrial-strength production deployment of deep learning models is beyond the scope of this book. Instead, in this chapter, you will learn how to create simple, self-contained deployments that...