Question: DeepLearning:-Optimizer(GradientDescent()) argument?

which kind of input is DeepLearning:-Optimizer(GradientDescent()) suppose to take? Nummeric value or function?

Please Wait...