MNIST Dataset
I will make a short introduction to the MNIST dataset. Feel free to skip it if you are already familiar with it.
MNIST is a dataset of images of handwritten digits. The images have a size of 28x28 and there are 60k images in the training set and 10k images in the test set. The images are grayscale and the values are between 0 and 255. The labels are integers between 0 and 9.
The following image shows some examples of the images in the dataset.
This dataset is very popular for tutorials because with the current computing power we can train a model that is very accurate in a short amount of time. The dataset is also very small so it is easy to download and use.