Struct engram::nn::Sequential

pub struct Sequential {
    pub layers: Vec<Box<dyn Layer>>,
    /* private fields */
}

Fields

layers: Vec<Box<dyn Layer>>

Implementations

impl Sequential

pub fn new() -> Self

pub fn add_layer<T>(&mut self, layer: T)

where T: Layer + 'static,

pub fn add_layers<T>(&mut self, layers: Vec<T>)

where T: Layer + 'static,

pub fn set_optimizer<T>(&mut self, optimizer: T)

where T: Optimizer + 'static,

pub fn set_regularization<T>(&mut self, regularization: T)

where T: Regularization + 'static,

pub fn train(&mut self)

pub fn eval(&mut self)

pub fn forward(&mut self, input: &Tensor) -> Tensor

pub fn backward(&mut self, target: &Tensor, loss_fn: &Loss) -> f64

pub fn fit( &mut self, dataset: &Dataset, loss_fn: &Loss, epochs: usize, batch_size: usize )

pub fn regularize(&mut self) -> f64

pub fn step(&mut self)

pub fn predict(&mut self, input: &Tensor) -> Tensor

Trait Implementations

impl Debug for Sequential

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

Formats the value using the given formatter.