how to find the filter size, kernels, stride and padding for the next convolutional layer in a neural network

I want to know how one can find the kernels, filter size in a neural network (say it is Alexnet). How we got filter size 11,11 and kernel 96 in first convolutional layer, then 5,5 and kernel 256 in second layer, also how to find max pooling, stride etc. Is there any formula to get these calculations or it is done with some other way? plz tell by example how to do it?