News

Some optimization algorithms also adapt the learning rates of the model parameters by looking at the gradient history (AdaGrad, RMSProp, and Adam).