Input Augmentation

What does it mean to “apply data augmentation to the inputs?” What are the inputs? Are the inputs the ‘alpaca’ and ‘not alpaca’ images? Do I need to iterate over the images in the train_dataset and validation_dataset variables and call the data_augmenter() function on each image?

It’s usually a good idea to identify your question by course number, week number, and assignment number.

I think what it means is:
x = data_augmentation(inputs)

… or something like that. I don’t recall what the left-side variable should be.

Don’t make it too complicated.

All of the iteration is buried out-of-sight later, in the cell that runs “model2.compile(…)”.

Thanks. Just so you know, I’m DEFINITELY not trying to complicate the assignment. It’s just that the keras documentation is complete garbage! If I wrote documentation in my job at the same level as the keras documentation, I would be immediately fired!!!

I agree.

The Keras documentation was written by people who were already TensorFlow experts, and already know too much about the topic to write good documentation.