Struct Tensor 

pub struct Tensor {
    pub data: Array1<f32>,
    pub grad: Option<Array1<f32>>,
    pub requires_grad: bool,
}

A 1-D tensor with optional gradient storage.

Fields

data: Array1<f32>

Forward-pass data.

grad: Option<Array1<f32>>

Accumulated gradient (filled by GradientTape::backward).

requires_grad: bool

Whether gradient computation is required for this tensor.

Implementations

impl Tensor

pub fn new(data: Array1<f32>) -> Self

Create a tensor that participates in gradient computation.

pub fn no_grad(data: Array1<f32>) -> Self

Create a tensor that does NOT participate in gradient computation.

Trait Implementations

impl Clone for Tensor

fn clone(&self) -> Tensor

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Tensor

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

Formats the value using the given formatter.