Google Code-in

Swift for TensorFlow Datasets:

As you would have known by now, Swift for TensorFlow Models is a collection of TensorFlow models written in Swift.

In Task 4, we asked you to train a simple Image Classification model on CIFAR-10 Dataset. As a Bonus subtask, you could/would have done Image Classification on the ImageNet Dataset. As mentioned there, there exists a subset of ImageNet which was created by Jeremy Howard known as Imagenette.

Your task is to code a script which can directly import the 160-px dataset and enable using the dataset within Colab or in the local environment.

You can use the template of the CIFAR10 dataset as an example of how to make this work.