I want to use SGD optimizer in tf.keras. But SGD detail said "Gradient descent (with momentum) optimizer.". Dose it mean SGD doesn\'t support "Randomly sh
