XnnpackOptions class

Configuration options for the XNNPACK execution provider.

XNNPACK is a highly optimized library for neural network inference on ARM and x86 CPUs. It uses SIMD instructions (NEON on ARM, SSE/AVX on x86) for maximum throughput.

OrtSessionWrapper.createWithProviders(
  modelPath,
  providers: [OrtProvider.xnnpack, OrtProvider.cpu],
  providerOptions: {
    OrtProvider.xnnpack: XnnpackOptions(numThreads: 4).toMap(),
  },
);

Constructors

XnnpackOptions({int numThreads = 0})

const

XnnpackOptions.fromMap(Map<String, String> map)

Creates XnnpackOptions from a provider options map.

factory

Properties

hashCode → int

The hash code for this object.

no setter, inherited

numThreads → int

Number of threads for XNNPACK's internal thread pool.

final

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

Methods

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

toMap() → Map<String, String>

Converts to the key-value map expected by the ONNX Runtime generic API.

toString() → String

A string representation of this object.

override

Operators

operator ==(Object other) → bool

The equality operator.

inherited