Enum NdArrayTensor 

pub enum NdArrayTensor {
    F64(NdArrayStorage<f64>),
    F32(NdArrayStorage<f32>),
    I64(NdArrayStorage<i64>),
    I32(NdArrayStorage<i32>),
    I16(NdArrayStorage<i16>),
    I8(NdArrayStorage<i8>),
    U64(NdArrayStorage<u64>),
    U32(NdArrayStorage<u32>),
    U16(NdArrayStorage<u16>),
    U8(NdArrayStorage<u8>),
    Bool(NdArrayStorage<bool>),
}

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

Tensor primitive used by the ndarray backend.

Supports both owned and borrowed (zero-copy) data via NdArrayStorage. When data is borrowed from external sources (like burnpack files), it remains zero-copy until a mutating operation is performed.

Variants

F64(NdArrayStorage<f64>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

F32(NdArrayStorage<f32>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

I64(NdArrayStorage<i64>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

I32(NdArrayStorage<i32>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

I16(NdArrayStorage<i16>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

I8(NdArrayStorage<i8>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

U64(NdArrayStorage<u64>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

U32(NdArrayStorage<u32>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

U16(NdArrayStorage<u16>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

U8(NdArrayStorage<u8>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

Bool(NdArrayStorage<bool>)

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

Implementations

impl NdArrayTensor

pub fn is_borrowed(&self) -> bool

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

Returns true if this tensor uses borrowed (zero-copy) storage.

impl NdArrayTensor

pub fn from_data(data: TensorData) -> NdArrayTensor

👎Deprecated since 0.22.0:

burn-ndarray is deprecated and will be removed in a future release. Use burn-flex for pure-Rust CPU execution (std, no_std, WebAssembly), or one of the CubeCL backends (burn-cuda, burn-rocm, burn-wgpu, burn-cpu) for GPU acceleration.

Create a new ndarray tensor from data.

This method attempts zero-copy loading when possible. If the data has properly aligned bytes that can be borrowed, it creates a borrowed tensor. Otherwise, it falls back to copying the data.

Zero-copy loading works when:

• The data’s bytes are properly aligned for the element type
• The bytes can be borrowed (e.g., from mmap’d file or static data)

Trait Implementations

impl Clone for NdArrayTensor

fn clone(&self) -> NdArrayTensor

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for NdArrayTensor

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

Formats the value using the given formatter.

impl From<ArrayBase<OwnedArcRepr<bool>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<bool>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<f32>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<f32>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<f64>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<f64>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<i8>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<i8>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<i16>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<i16>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<i32>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<i32>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<i64>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<i64>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<u8>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<u8>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<u16>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<u16>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<u32>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<u32>) -> NdArrayTensor

Converts to this type from the input type.

impl From<ArrayBase<OwnedArcRepr<u64>, Dim<IxDynImpl>>> for NdArrayTensor

fn from(value: SharedArray<u64>) -> NdArrayTensor

Converts to this type from the input type.

impl TensorMetadata for NdArrayTensor

type Device = NdArrayDevice

The device type associated with the tensor.

fn dtype(&self) -> DType

Get the dtype of the tensor.

fn shape(&self) -> Shape

Get the shape of the tensor.

fn rank(&self) -> usize

Get the number of dimensions of the tensor.

fn device(&self) -> NdArrayDevice

Get the device associated with the tensor.

fn can_mut(&self) -> bool

Whether the tensor’s buffer can be mutated in place — i.e. this handle uniquely owns it, so an in-place op (slice_assign, an inplace kernel) writes the existing allocation instead of copying it first.

fn scheme(&self) -> QuantScheme

Get the quantization scheme for a quantized float tensor.