Dataset of 50,000 32x32 color training images, labeled over 10 categories, and 10,000 test images.
Value
Lists of training and test data: train$x, train$y, test$x, test$y.
The x data is an array of RGB image data with shape (num_samples, 3, 32,
32).
The y data is an array of category labels (integers in range 0-9) with
shape (num_samples).
See also
Other datasets: dataset_boston_housing() dataset_cifar100() dataset_fashion_mnist() dataset_imdb() dataset_mnist() dataset_reuters()