{# Copyright (c) 2015, NVIDIA CORPORATION. All rights reserved. #}
You should enter a train_val network here. Some cleanup is done for you automatically:
You should enter a Torch network description here. Refer to the documentation for more information.