Trait autograd::ops::Op

pub trait Op {
    fn name(&self) -> &str;
    fn grad(
        &self, 
        gy: &Tensor, 
        inputs: &[&Tensor], 
        output: &Tensor
    ) -> Vec<Option<Tensor>>;
    fn compute(&mut self, xs: &[&NdArray], train: bool) -> NdArray;
}

Represents a operation node in a computation graph. Tensor wraps trait-object of this.

Required Methods

fn name(&self) -> &str

Name of this op

fn grad(
    &self,
    gy: &Tensor,
    inputs: &[&Tensor],
    output: &Tensor
) -> Vec<Option<Tensor>>

Returns gradient for each input node by use of output gradient.

Arguments

• gy - Gradient of output of this op
• inputs - Tensor level representation of compute::xs
• output - Tensor level representation of compute's return value

fn compute(&mut self, xs: &[&NdArray], train: bool) -> NdArray

Actually runs this op. num of inputs : N, num of outputs: 1

Implementors

• impl Op for DummyOp
• impl Op for RandomNormal
• impl Op for RandomUniform
• impl Op for StandardNormal
• impl Op for StandardUniform
• impl Op for Bernoulli
• impl Op for Exponential
• impl Op for LogNormal
• impl Op for Gamma