How can we get the best deep learning model from the multiple deep learning models?

Hello Team,

I'm implementing an image classification use-case in dataiku using multiple deep learning models such as CNN, RESNET, DENSENET and etc.,. and want to get champion model out of it. Is there any out of box capability that gives champion model out of tried deep learning models based on chosen metric? I'm aware that it is possible for basic statistical models wondering how can we achieve same for deep learning models out of box?

Thanks & Regards,

Lakshmi Narayana

Operating system used: Linux(CentOS)