Enum TensorLayout 

pub enum TensorLayout {
    Nchw,
    Nhwc,
    Ncdhw,
    Ndhwc,
    RowMajor,
}

Memory layout convention for multi-dimensional tensors.

The layout determines how logical indices map to linear memory offsets. NHWC layouts are generally preferred on modern NVIDIA GPUs because they enable Tensor Core utilisation, while NCHW is the traditional PyTorch default.

Variants

Nchw

Batch, Channels, Height, Width – PyTorch default.

Nhwc

Batch, Height, Width, Channels – Tensor Core optimal.

Ncdhw

Batch, Channels, Depth, Height, Width – 3-D volumetric.

Ndhwc

Batch, Depth, Height, Width, Channels – 3-D channels-last.

RowMajor

Generic row-major layout for 2-D tensors (matrices) and MoE intermediates.

Implementations

impl TensorLayout

pub const fn spatial_dims(self) -> usize

Returns the number of spatial dimensions implied by this layout.

pub const fn expected_ndim(self) -> usize

Returns the expected number of tensor dimensions (including N and C).

pub const fn is_channels_last(self) -> bool

Returns true if this layout places channels last (NHWC or NDHWC).

Trait Implementations

impl Clone for TensorLayout

fn clone(&self) -> TensorLayout

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for TensorLayout

impl Debug for TensorLayout

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

Formats the value using the given formatter.

impl Eq for TensorLayout

impl Hash for TensorLayout

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for TensorLayout

fn eq(&self, other: &TensorLayout) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for TensorLayout