Enum Layer 

pub enum Layer {
    Linear(AffFunc),
    ReLU(usize),
    LeakyReLU(usize, f64),
    HardTanh(usize),
    HardSigmoid(usize),
    Argmax,
    ClassChar(usize),
}

A simple enum type to conveniently specify the layer structure of a neural network. Each Layer corresponds to one piece-wise linear function.

Variants

Linear(AffFunc)

A fully connected linear layer

ReLU(usize)

The ReLU applied to the i-th component of the input

LeakyReLU(usize, f64)

The Leaky ReLU applied to the i-th component of the input

HardTanh(usize)

The hard hyperbolic tangent applied to the i-th component of the input

HardSigmoid(usize)

The hard sigmoid applied to the i-th component of the input

Argmax

The argmax function

ClassChar(usize)

A binary version of the argmax called class characterization. The result is a boolean indicating whether the input belongs to the specified class or not.

Trait Implementations

impl Clone for Layer

fn clone(&self) -> Layer

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for Layer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.