Enum OrtExecutionProvider 

pub enum OrtExecutionProvider {
    CPU,
    CUDA {
        device_id: Option<i32>,
        gpu_mem_limit: Option<usize>,
        arena_extend_strategy: Option<String>,
        cudnn_conv_algo_search: Option<String>,
        do_copy_in_default_stream: Option<bool>,
        cudnn_conv_use_max_workspace: Option<bool>,
    },
    DirectML {
        device_id: Option<i32>,
    },
    OpenVINO {
        device_type: Option<String>,
        num_threads: Option<usize>,
    },
    TensorRT {
        device_id: Option<i32>,
        max_workspace_size: Option<usize>,
        max_batch_size: Option<usize>,
        min_subgraph_size: Option<usize>,
        fp16_enable: Option<bool>,
    },
    CoreML {
        ane_only: Option<bool>,
        subgraphs: Option<bool>,
    },
    WebGPU,
}

Execution providers for ONNX Runtime.

This enum represents the different execution providers that can be used with ONNX Runtime for model inference.

Variants

CPU

CPU execution provider (always available)

CUDA

NVIDIA CUDA execution provider

Fields

device_id: Option<i32>

CUDA device ID (default: 0)

gpu_mem_limit: Option<usize>

Memory limit in bytes (optional)

arena_extend_strategy: Option<String>

Whether to use arena allocator (default: true)

cudnn_conv_algo_search: Option<String>

CUDNN convolution algorithm search (default: “EXHAUSTIVE”)

do_copy_in_default_stream: Option<bool>

Whether to do copy in default stream (default: true)

cudnn_conv_use_max_workspace: Option<bool>

CUDNN convolution use max workspace (default: true)

DirectML

DirectML execution provider (Windows only)

Fields

device_id: Option<i32>

DirectML device ID (default: 0)

OpenVINO

OpenVINO execution provider

Fields

device_type: Option<String>

Device type (e.g., “CPU”, “GPU”, “MYRIAD”)

num_threads: Option<usize>

Number of threads (optional)

TensorRT

TensorRT execution provider

Fields

device_id: Option<i32>

TensorRT device ID (default: 0)

max_workspace_size: Option<usize>

Maximum workspace size in bytes

max_batch_size: Option<usize>

Maximum batch size

min_subgraph_size: Option<usize>

Minimum subgraph size

fp16_enable: Option<bool>

FP16 enable flag

CoreML

CoreML execution provider (macOS/iOS only)

Fields

ane_only: Option<bool>

Use Apple Neural Engine only

subgraphs: Option<bool>

Enable subgraphs

WebGPU

WebGPU execution provider

Trait Implementations

impl Clone for OrtExecutionProvider

fn clone(&self) -> OrtExecutionProvider

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for OrtExecutionProvider

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

Formats the value using the given formatter.

impl Default for OrtExecutionProvider

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for OrtExecutionProvider

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for OrtExecutionProvider

fn eq(&self, other: &OrtExecutionProvider) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for OrtExecutionProvider

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for OrtExecutionProvider