Struct NdArrayQTensor 

pub struct NdArrayQTensor {
    pub qtensor: NdArrayTensor,
    pub scheme: QuantScheme,
    pub qparams: Vec<f32>,
    pub global: Option<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.

A quantized tensor for the ndarray backend.

Fields

qtensor: 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.

The quantized tensor.

scheme: QuantScheme

👎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.

The quantization scheme.

qparams: Vec<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.

The block scales.

global: Option<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.

The per-tensor scale that qparams are expressed relative to, for a two-level scheme.

Implementations

impl NdArrayQTensor

pub fn strategy(&self) -> QuantizationStrategy

👎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 the quantization strategy, including quantization parameters, for the given tensor.

Trait Implementations

impl Clone for NdArrayQTensor

fn clone(&self) -> NdArrayQTensor

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for NdArrayQTensor

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

Formats the value using the given formatter.

impl TensorMetadata for NdArrayQTensor

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) -> Self::Device

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.