Crate axonml_tensor 

Axonml Tensor - N-Dimensional Array for Machine Learning

This crate provides the core Tensor type that serves as the foundation for all machine learning operations in Axonml. Tensors are multi-dimensional arrays with support for automatic broadcasting, device placement, and efficient memory sharing through views.

Key Features

• N-dimensional tensor with arbitrary shape
• Automatic broadcasting for operations
• Efficient views and slicing (zero-copy where possible)
• Device-agnostic (CPU, CUDA, Vulkan, etc.)
• Generic over data type (f32, f64, i32, etc.)

Example

use axonml_tensor::{zeros, ones, Tensor};

// Create tensors using factory functions
let a = zeros::<f32>(&[2, 3]);
let b = ones::<f32>(&[2, 3]);

// Arithmetic operations
let c = a.add(&b).unwrap();
let d = c.mul_scalar(2.0);

// Reductions
let sum = d.sum();
let mean = d.mean();

@version 0.1.0 @author AutomataNexus Development Team

Re-exports

pub use shape::Shape;

pub use shape::Strides;

pub use tensor::Tensor;

pub use creation::*;

Modules

creation

Tensor Creation Functions

ops

Tensor Operations - Mathematical and Structural Operations

prelude

Convenient imports for common usage.

shape

Shape and Strides - Tensor Dimension Management

tensor

Tensor - Core N-Dimensional Array Type

view

Views and Slicing - Tensor Indexing Operations

Enums

DType

Runtime representation of tensor data types.

Device

Represents a compute device where tensors can be allocated and operations executed.

Error

The main error type for Axonml operations.

Type Aliases

Result

A specialized Result type for Axonml operations.