Deploying a fastai model trained on an image dataset

In the Deploying a fastai model trained on a tabular dataset recipe, we went through the process of deploying a model trained on a tabular dataset. We deployed a model that predicted whether an individual would have an income over 50,000 based on a set of characteristics called scoring parameters, including education, job category, and hours worked per week. To do this deployment, we needed a way to allow the user to select values for the scoring parameters and then show the prediction made by the trained fastai model on these scoring parameters.

In this recipe, we will deploy the image classification model that you trained in the Training a classification model with a standalone vision dataset recipe of Chapter 6, Training Models with Visual Data. This model predicts what fruit or vegetable is depicted in an image. Unlike the deployment of the tabular dataset model, to deploy the image dataset model we will need to be able...