ForwardDiff

nncombinator::layer

Trait ForwardDiff 

Source 
pub trait ForwardDiff<U>: PreTrain<U>
where
    U: UnitValue<U>,
{
    // Required method
    fn forward_diff(
        &self,
        input: Self::Input,
    ) -> Result<Self::OutStack, EvaluateError>;
}
Expand description

Trait that defines the function of differential application of inputs in the process of forward propagation to neural networks.

Required Methods§

Source

fn forward_diff( &self, input: Self::Input, ) -> Result<Self::OutStack, EvaluateError>

Forward propagation (differential application)

§Arguments
  • input - input
§Errors

This function may return the following errors

Implementors§

Source§

impl<T, U> ForwardDiff<U> for T
where T: PreTrain<U> + Sized, U: UnitValue<U>,