System.Numerics.Tensors 10.0.0-rc.2.25502.107

About

Provides methods for performing mathematical operations over tensors. This library offers both high-level tensor types and low-level primitives for working with multi-dimensional numeric data. Many operations are accelerated to use SIMD (Single instruction, multiple data) operations supported by the CPU where available.

Key Features

• High-level tensor types: Tensor<T>, TensorSpan<T>, ReadOnlyTensorSpan<T> for working with multi-dimensional arrays
• Low-level tensor primitives: TensorPrimitives for efficient span-based operations
• Generic support for various numeric types (float, double, int, etc.)
• Element-wise arithmetic: Add, Subtract, Multiply, Divide, Exp, Log, Cosh, Tanh, etc.
• Tensor arithmetic: CosineSimilarity, Distance, Dot, Normalize, Softmax, Sigmoid, etc.
• SIMD-accelerated operations for improved performance

How to Use

using System.Numerics.Tensors;

var movies = new[] {
    new { Title="The Lion King", Embedding= new [] { 0.10022575f, -0.23998135f } },
    new { Title="Inception", Embedding= new [] { 0.10327095f, 0.2563685f } },
    new { Title="Toy Story", Embedding= new [] { 0.095857024f, -0.201278f } },
    new { Title="Pulp Function", Embedding= new [] { 0.106827796f, 0.21676421f } },
    new { Title="Shrek", Embedding= new [] { 0.09568083f, -0.21177962f } }
};
var queryEmbedding = new[] { 0.12217915f, -0.034832448f };

// Using TensorPrimitives for low-level span operations
var top3MoviesTensorPrimitives =
    movies
        .Select(movie =>
            (
                movie.Title,
                Similarity: TensorPrimitives.CosineSimilarity(queryEmbedding, movie.Embedding)
            ))
        .OrderByDescending(movies => movies.Similarity)
        .Take(3);

foreach (var movie in top3MoviesTensorPrimitives)
{
    Console.WriteLine(movie);
}

// Using higher-level Tensor types for multi-dimensional operations
float[] data1 = [1f, 2f, 3f, 4f, 5f, 6f];
float[] data2 = [6f, 5f, 4f, 3f, 2f, 1f];
Tensor<float> tensor1 = Tensor.Create(data1, [2, 3]); // 2x3 tensor
Tensor<float> tensor2 = Tensor.Create(data2, [2, 3]); // 2x3 tensor
Tensor<float> result = tensor1 + tensor2;

Main Types

The main types provided by this library are:

• System.Numerics.Tensors.TensorPrimitives - Low-level operations on spans of numeric data
• System.Numerics.Tensors.Tensor<T> - Generic tensor class for multi-dimensional arrays
• System.Numerics.Tensors.TensorSpan<T> - Span-like view over tensor data
• System.Numerics.Tensors.ReadOnlyTensorSpan<T> - Read-only span-like view over tensor data
• System.Numerics.Tensors.Tensor - Static class with high-level tensor operations

Additional Documentation

• API documentation

Feedback & Contributing

System.Numerics.Tensors is released as open source under the MIT license. Bug reports and contributions are welcome at the GitHub repository.