So i have a DL model which has been implemented in both tensorflow and pytorch. Now since both implementations are in the same package, i do not want any user to have to ins
