A hyperparameter that controls the step size at each iteration while moving toward a minimum of a loss function during training.
Last updated 11 months ago