Model parameters constitute models to encode data patterns while hyperparameters control how the models are training. The latter are set manually to tune the model training.