Trait gad::net_ext::SingleOutputNet

pub trait SingleOutputNet<Data, Algebra>: Net<Algebra> where
    Algebra: HasGradientReader + CoreAlgebra<Data, Value = Self::Output>, {
    fn add_square_loss(self) -> SquareLoss<Self, Data>
    where
        Self: Sized,
    { ... }
}

Extensions trait when the network has a single output value (e.g. a single multi-dimensional array).

Provided methods

fn add_square_loss(self) -> SquareLoss<Self, Data> where
    Self: Sized, 

A network that takes an additional input and returns the L2-distance with the output of the initial network.

Loading content...

Implementors

impl<Data, Algebra, N> SingleOutputNet<Data, Algebra> for N where
    N: Net<Algebra>,
    Algebra: HasGradientReader + CoreAlgebra<Data, Value = Self::Output>, 

Loading content...