IOnnxModel<T>

Interfaces · AiDotNet.Interfaces

Defines the contract for ONNX model wrappers that provide cross-platform model inference.

Implements: IDisposable

Known implementations: OnnxModel<T>

For Beginners

ONNX (Open Neural Network Exchange) is a universal format for neural network models. This interface allows you to:

  • Load models trained in PyTorch, TensorFlow, or other frameworks
  • Run inference using CPU, GPU (CUDA), TensorRT, or DirectML
  • Convert between AiDotNet tensors and ONNX tensors automatically

How It Works

This interface provides a unified way to work with ONNX models in AiDotNet. It supports loading models from files, byte arrays, or URLs, and provides methods for running inference with AiDotNet Tensor types.

Properties

Property Summary
IOnnxModelMetadata Metadata { get; } Gets the metadata about the loaded ONNX model.
bool IsLoaded { get; } Gets whether the model has been successfully loaded and is ready for inference.
string ExecutionProvider { get; } Gets the execution provider currently being used (CPU, CUDA, TensorRT, DirectML).

Methods

Method Summary
IReadOnlyDictionary<string, Tensor<T>> Run(IReadOnlyDictionary<string, Tensor<T>> inputs) Runs inference with named inputs.
Task WarmUpAsync(CancellationToken cancellationToken) Warms up the model asynchronously.
Task<IReadOnlyDictionary<string, Tensor<T>>> RunAsync(IReadOnlyDictionary<string, Tensor<T>> inputs, CancellationToken cancellationToken) Runs inference asynchronously with named inputs.
Task<Tensor<T>> RunAsync(Tensor<T> input, CancellationToken cancellationToken) Runs inference asynchronously with named inputs.
Tensor<T> Run(Tensor<T> input) Runs inference with named inputs.
void WarmUp() Warms up the model by running a single inference with dummy data.