Enum TensorOutput 

pub enum TensorOutput {
    F32 {
        shape: Vec<i64>,
        data: Vec<f32>,
    },
    I64 {
        shape: Vec<i64>,
        data: Vec<i64>,
    },
}

Generic tensor output supporting multiple data types.

This enum represents the raw output from an inference engine without making assumptions about the semantic meaning of the data. It’s the responsibility of the caller (model layer) to interpret and validate the output shape and type.

Variants

F32

32-bit floating point tensor

Fields

shape: Vec<i64>

data: Vec<f32>

I64

64-bit integer tensor (commonly used for token IDs)

Fields

shape: Vec<i64>

data: Vec<i64>

Implementations

impl TensorOutput

pub fn shape(&self) -> &[i64]

Returns the shape of the tensor.

pub fn ndim(&self) -> usize

Returns the number of dimensions.

pub fn len(&self) -> usize

Returns the total number of elements.

pub fn is_empty(&self) -> bool

Returns true if the tensor has no elements.

pub fn try_into_array2_f32( self, ) -> Result<ArrayBase<OwnedRepr<f32>, Dim<[usize; 2]>>, OCRError>

Attempts to extract as a 2D f32 array.

Errors

Returns an error if:

• The tensor is not f32 type
• The tensor is not 2-dimensional
• The data length doesn’t match the shape

pub fn try_into_array3_f32( self, ) -> Result<ArrayBase<OwnedRepr<f32>, Dim<[usize; 3]>>, OCRError>

Attempts to extract as a 3D f32 array.

pub fn try_into_array4_f32( self, ) -> Result<ArrayBase<OwnedRepr<f32>, Dim<[usize; 4]>>, OCRError>

Attempts to extract as a 4D f32 array.

pub fn try_into_array_f32( self, ) -> Result<ArrayBase<OwnedRepr<f32>, Dim<IxDynImpl>>, OCRError>

Attempts to extract as a dynamic-dimensional f32 array.

pub fn try_into_array2_i64( self, ) -> Result<ArrayBase<OwnedRepr<i64>, Dim<[usize; 2]>>, OCRError>

Attempts to extract as a 2D i64 array.

pub fn try_into_array3_i64( self, ) -> Result<ArrayBase<OwnedRepr<i64>, Dim<[usize; 3]>>, OCRError>

Attempts to extract as a 3D i64 array.

Trait Implementations

impl Clone for TensorOutput

fn clone(&self) -> TensorOutput

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TensorOutput

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

Formats the value using the given formatter.