Training a recommender system on a small curated dataset

You may recall that Chapter 1, Getting Started with fastai, described applications supported by fastai to cover four types of datasets: tabular, text, recommender systems, and images. In Chapter 2, Exploring and Cleaning Up Data with fastai, you saw sections on examining tabular datasets, text datasets, and image datasets.

You may have wondered why there wasn't a section on examining recommender system datasets. The reason is that the data ingestion process for recommender systems in fastai is identical to the process for tabular datasets, as you will see in this section. While the ingestion process for recommender systems in fastai is identical to the ingestion process for tabular datasets, fastai does provide model training details that are specifically intended for recommender systems.

In this section, we will go through the process of training a recommender system on a curated dataset to learn how to train recommender...