Two output nodes in uff model

Can I have two output nodes in uff model and use tensorRT to run it?