The CNN example diagram shows 3 conv & pooling layers but the implementation only has 2

[grantgasser]

In the CNN tutorial, the diagram shows a CNN with 3 convolutional and pooling layers, but the implementation has 2 convolutional and pooling layers, which may be confusing or misleading to beginners.

I'd be happy to add the 3rd layer to the implementation and make a PR.

Cheers.