org.tensorflow.op.annotation

Annotation Type Operator

@Documented
 @Target(value=TYPE)
 @Retention(value=SOURCE)
public @interface Operator

Annotation used by classes to make TensorFlow operations conveniently accessible via org.tensorflow.op.Ops.

An annotation processor (org.tensorflow.processor.OperatorProcessor) builds the Ops class by aggregating all classes annotated as @Operators. Each annotated class must have at least one public static factory method named create that accepts a Scope as its first argument. The processor then adds a convenience method in the Ops class. For example:

{@code

Optional Element Summary

Optional Elements

Modifier and Type	Optional Element and Description
String	group Specify an optional group within the Ops class.
String	name Name for the wrapper method used in the Ops class.

Element Detail

group

public abstract String group

Specify an optional group within the Ops class.

By default, an annotation processor will create convenience methods directly in the Ops class. An annotated operator may optionally choose to place the method within a group. For example:

{@code

Default:

""

name

public abstract String name

Name for the wrapper method used in the Ops class.

By default, a processor derives the method name in the Ops class from the class name of the operator. This attribute allow you to provide a different name instead. For example:

{@code

Default:

""