Find out the input and output layers

how do I know, how many input and output layers are in deep speech?