Difference between MLS and DLS for the input matrix X and output matrix Y

I notice that the input matrix X, W, Y to neural network from the DLS are inverted to the matrix X,W,Y from MLS.
I prefer the matrix setup way from MLS, it seems to be easier to follow.
Can anyone advise?

There is no agreed standard in this regard. You may find any possible arrangement of the dimensions, depending on the author of the dataset.

Thank you for the advice.
I wrap my head around just for the dimensions change between MLS and DLS.

You’re pretty much always going to have to inspect the data set before you use it, to see how it’s organized.