winml.h header

This header is used by Windows Machine Learning. For more information, see:

• Windows Machine Learning

winml.h contains the following programming interfaces:

Interfaces

IWinMLEvaluationContext Represents the context to bind inputs and outputs to a WinML model.
IWinMLModel Represents a Windows Machine Learning model with corresponding metadata; includes model descriptions (name, author, versioning, etc.), as well as expected inputs and outputs.
IWinMLRuntime Represents the runtime to load and evaluate a WinML model.
IWinMLRuntimeFactory Represents the factory that creates the WinML runtime for model loading and evaluation.

Structures

WINML_BINDING_DESC Contains a description of the WinML binding.
WINML_IMAGE_BINDING_DESC Contains properties for the binding of the type image.
WINML_IMAGE_VARIABLE_DESC Contains properties for the image variable description.
WINML_MAP_BINDING_DESC Contains properties for the binding of type map.
WINML_MAP_VARIABLE_DESC Contains description properties of the map variable.
WINML_MODEL_DESC Contains description properties of the model.
WINML_RESOURCE_BINDING_DESC Contains description properties of the resource binding.
WINML_SEQUENCE_BINDING_DESC Contains description properties of the sequence binding.
WINML_SEQUENCE_VARIABLE_DESC Contains description properties of the sequence variable.
WINML_TENSOR_BINDING_DESC Contains description properties of the tensor binding.
WINML_TENSOR_VARIABLE_DESC Contains description properties of the tensor variable.
WINML_VARIABLE_DESC Contains description properties of the variable.

Enumerations

WINML_BINDING_TYPE Specifies the different types of WinML bindings.
WINML_FEATURE_TYPE Specifies the different types of WinML features.
WINML_RUNTIME_TYPE Specifies the different types of WinML runtimes.
WINML_TENSOR_DATA_TYPE Specifies the different data types of WinML tensors.