This reference architecture guide describes the configuration and methodology of performing DL processing on a complex dataset using easy to use software tools and specialized hardware configurations designed to show the advantages in processing speeds and increased productivity of this type of compute-intensive task.