So far, we have been working with a subset of the MNIST dataset. This subset contains only 10,000 rows, but the official MNIST dataset actually contains 60,000 images for training and 10,000 images for testing.
So let’s download the full dataset and see how our app holds up.
...
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