PyTorch C++ FrontEnd returning multiple Tensors

I was wonder how can I return a std::vectortorch::Tensor in my forward pass of a Module Class,
I read about the Macro of FORWARD_HAS_DEFAULT_ARGS in the docs, but didn’t really
understand how to use it, and also how to use it for making it possible to return a vector in return.
Thank you in advance.

I am also wondering how is it possible to do this.
I was thinking about creating a strcuture like

typedef struct
  at::Tensor index;
	at::Tensor mask;
} Ret;

and then define the function as

Ret *function(args)

instead of

at::Tensor function(args)

I wish to use this c++ implementation with a python wrapper for pytorch. How can I reflect this modification at::Tensor to Ret * in PyTorch implementation?