The MNIST dataset is a great one to use when starting to learn about federated learning. MNIST can be thought of as the “Hello, World!” of machine learning, and is mainly used to experiment with machine learning algorithms and to compare their relative strengths.
The dataset is a collection of handwritten digits that can be used to train and test different supervised machine learning algorithms. The data is split into two subsets, with 60,000 images belonging to the training set and 10,000 images belonging to the testing set, with a level of variety to give robustness to an appropriately trained model.