org.eclipse.kura.ai.inference

Class TensorDescriptor

java.lang.Object

org.eclipse.kura.ai.inference.TensorDescriptor

@ProviderType
public class TensorDescriptor
extends java.lang.Object

The TensorDescriptor class describes the Tensor used as input or output of a model for Artificial Intelligence and Machine Learning algorithms

Since:

2.3

Constructor Summary

Constructors

Constructor and Description
TensorDescriptor(java.lang.String name, java.lang.String type, java.util.Optional<java.lang.String> format, java.util.List<java.lang.Long> shape, java.util.Map<java.lang.String,java.lang.Object> parameters) Instantiates a tensor descriptor

Method Summary

Modifier and Type	Method and Description
static TensorDescriptorBuilder	builder(java.lang.String name, java.lang.String type, java.util.List<java.lang.Long> shape) Instantiates a builder for a TensorDescriptor
java.util.Optional<java.lang.String>	getFormat() Return the format of the data.
java.lang.String	getName() Return the name of the tensor
java.util.Map<java.lang.String,java.lang.Object>	getParameters() Return the optional parameters assign to the tensor
java.util.List<java.lang.Long>	getShape() Return the shape of the data as the size of a multi-dimensional matrix.
java.lang.String	getType() Return the type of data contained in the tensor.
TensorDescriptorBuilder	toBuilder() Creates a new TensorDescriptorBuilder and initialises it from this TensorDescriptor instance.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TensorDescriptor

public TensorDescriptor(java.lang.String name,
                        java.lang.String type,
                        java.util.Optional<java.lang.String> format,
                        java.util.List<java.lang.Long> shape,
                        java.util.Map<java.lang.String,java.lang.Object> parameters)

Instantiates a tensor descriptor

Parameters:

name - the name of the tensor

type - a string representing the type of data contained in the tensor. Its value is implementation specific, so a user should refer to the implementation documentation to figure out the allowed values.

shape - the shape of the data

Method Detail

builder

public static TensorDescriptorBuilder builder(java.lang.String name,
                                              java.lang.String type,
                                              java.util.List<java.lang.Long> shape)

Instantiates a builder for a TensorDescriptor

Parameters:

name - the name of the tensor

type - a string representing the type of data contained in the tensor. Its value is implementation specific, so a user should refer to the implementation documentation to figure out the allowed values.

shape - the shape of the data as the size of a multi-dimensional matrix

toBuilder

public TensorDescriptorBuilder toBuilder()

Creates a new TensorDescriptorBuilder and initialises it from this TensorDescriptor instance.

Returns:

a new TensorDescriptorBuilder

getName

public java.lang.String getName()

Return the name of the tensor

Returns:

a string representing the tensor name

getType

public java.lang.String getType()

Return the type of data contained in the tensor. Its value is implementation specific, so a user should refer to the implementation documentation to figure out the allowed values.

Returns:

a string representing the type of data contained in the tensor

getFormat

public java.util.Optional<java.lang.String> getFormat()

Return the format of the data. It represents how the data are organised or grouped in the tensor. Its value is implementation specific, so a user should refer to the implementation documentation to figure out the allowed values.

Returns:

an optional string representing the format of the data in the tensor

getShape

public java.util.List<java.lang.Long> getShape()

Return the shape of the data as the size of a multi-dimensional matrix.

Returns:

an unmodifiable list of longs representing the shape of the data

getParameters

public java.util.Map<java.lang.String,java.lang.Object> getParameters()

Return the optional parameters assign to the tensor

Returns:

an unmodifiable map containing the tensor parameters