onnxruntime helpers¶

Frequent functions
Python Wrapper OrtDevice
C class, OrtDevice or C_OrtDevice
OrtMemoryInfo
C classes, frequent types
Rare functions

Frequent functions ¶

onnxruntime.get_device() → str¶: Return the device used to compute the prediction (CPU, MKL, …)

<<<

import onnxruntime
print(onnxruntime.get_device())

>>>

CPU

onnxruntime.get_all_providers() → List[str]¶: Return list of Execution Providers that this version of Onnxruntime can support. The order of elements represents the default priority order of Execution Providers from highest to lowest.

<<<

import pprint
import onnxruntime
pprint.pprint(onnxruntime.get_all_providers())

>>>

    ['TensorrtExecutionProvider',
     'CUDAExecutionProvider',
     'MIGraphXExecutionProvider',
     'ROCMExecutionProvider',
     'OpenVINOExecutionProvider',
     'DnnlExecutionProvider',
     'NupharExecutionProvider',
     'TvmExecutionProvider',
     'VitisAIExecutionProvider',
     'NnapiExecutionProvider',
     'CoreMLExecutionProvider',
     'ArmNNExecutionProvider',
     'ACLExecutionProvider',
     'DmlExecutionProvider',
     'RknpuExecutionProvider',
     'CPUExecutionProvider']

onnxruntime.get_available_providers() → List[str]¶: Return list of available Execution Providers in this installed version of Onnxruntime. The order of elements represents the default priority order of Execution Providers from highest to lowest.

<<<

import onnxruntime
import pprint
pprint.pprint(onnxruntime.get_available_providers())

>>>

    ['CPUExecutionProvider']

onnxruntime.set_default_logger_severity(arg0: int) → None¶: Sets the default logging severity. 0:Verbose, 1:Info, 2:Warning, 3:Error, 4:Fatal

onnxruntime.set_seed(arg0: int) → None¶: Sets the seed used for random number generation in Onnxruntime.

Python Wrapper OrtDevice ¶

class onnxruntime.OrtDevice(c_ort_device)[source]¶

A data structure that exposes the underlying C++ OrtDevice

device_id()[source]¶

device_type()[source]¶

static make(ort_device_name, device_id)[source]¶

C class, OrtDevice or C_OrtDevice ¶

class onnxruntime.capi._pybind_state.OrtDevice¶

ONNXRuntime device informaion.

static cpu() → int¶

static cuda() → int¶

static default_memory() → int¶

device_id(self: onnxruntime.capi.onnxruntime_pybind11_state.OrtDevice) → int¶: Device Id.

device_type(self: onnxruntime.capi.onnxruntime_pybind11_state.OrtDevice) → int¶: Device Type.

OrtMemoryInfo ¶

class onnxruntime.capi._pybind_state.OrtMemoryInfo¶

C classes, frequent types ¶

class onnxruntime.capi._pybind_state.ModelMetadata¶

Pre-defined and custom metadata about the model. It is usually used to identify the model used to run the prediction and facilitate the comparison.

property custom_metadata_map¶: additional metadata

property description¶: description of the model

property domain¶: ONNX domain

property graph_description¶: description of the graph hosted in the model

property graph_name¶: graph name

property producer_name¶: producer name

property version¶: version of the model

class onnxruntime.capi._pybind_state.OrtMemType¶

Members:

CPU_INPUT

CPU_OUTPUT

CPU

DEFAULT

property name¶

Rare functions ¶

onnxruntime.capi._pybind_state.clear_training_ep_instances() → None¶: Clean the execution provider instances used in ort training module.

onnxruntime.capi._pybind_state.create_and_register_allocator(arg0: OrtMemoryInfo, arg1: OrtArenaCfg) → None¶

onnxruntime.capi._pybind_state.enable_telemetry_events() → None¶: Enables platform-specific telemetry collection where applicable.

onnxruntime.capi._pybind_state.disable_telemetry_events() → None¶: Disables platform-specific telemetry collection.

Summary of onnxruntime API

OrtValue

onnxruntime helpers¶

Frequent functions¶

Python Wrapper OrtDevice¶

C class, OrtDevice or C_OrtDevice¶

OrtMemoryInfo¶

C classes, frequent types¶

Rare functions¶