Packages

c

org.tensorflow.metadata.v0

MetaOptimizationTarget

final class MetaOptimizationTarget extends GeneratedMessageV3 with MetaOptimizationTargetOrBuilder

The high-level objectives described by this problem statement. These
objectives provide a basis for ranking models and can be optimized by a meta
optimizer (e.g. a grid search over hyperparameters). A solution provider may
also directly use the meta optimization targets to heuristically select
losses, etc without any meta-optimization process. If not specified, the
high-level meta optimization target is inferred from the task. These
objectives do not need to be differentiable, as the solution provider may use
proxy function to optimize model weights. Target definitions include tasks,
metrics, and any weighted combination of them.

Protobuf type tensorflow.metadata.v0.MetaOptimizationTarget

Source
MetaOptimizationTarget.java
Linear Supertypes
MetaOptimizationTargetOrBuilder, GeneratedMessageV3, Serializable, AbstractMessage, Message, MessageOrBuilder, AbstractMessageLite[MessageType, BuilderType], MessageLite, MessageLiteOrBuilder, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. MetaOptimizationTarget
  2. MetaOptimizationTargetOrBuilder
  3. GeneratedMessageV3
  4. Serializable
  5. AbstractMessage
  6. Message
  7. MessageOrBuilder
  8. AbstractMessageLite
  9. MessageLite
  10. MessageLiteOrBuilder
  11. AnyRef
  12. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. Protected

Value Members

  1. def equals(obj: AnyRef): Boolean
    Definition Classes
    MetaOptimizationTarget → AbstractMessage → Message → AnyRef → Any
    Annotations
    @Override()
  2. def findInitializationErrors(): List[String]
    Definition Classes
    AbstractMessage → MessageOrBuilder
  3. def getAllFields(): Map[FieldDescriptor, AnyRef]
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  4. def getDefaultInstanceForType(): MetaOptimizationTarget
    Definition Classes
    MetaOptimizationTarget → MessageOrBuilder → MessageLiteOrBuilder
    Annotations
    @Override()
  5. def getDescriptorForType(): Descriptor
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  6. def getField(field: FieldDescriptor): AnyRef
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  7. def getInitializationErrorString(): String
    Definition Classes
    AbstractMessage → MessageOrBuilder
  8. def getObjectiveCombinationCase(): ObjectiveCombinationCase
    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
  9. def getOneofFieldDescriptor(oneof: OneofDescriptor): FieldDescriptor
    Definition Classes
    GeneratedMessageV3 → AbstractMessage → MessageOrBuilder
  10. def getParserForType(): Parser[MetaOptimizationTarget]
    Definition Classes
    MetaOptimizationTarget → GeneratedMessageV3 → Message → MessageLite
    Annotations
    @Override()
  11. def getPerformanceMetric(): PerformanceMetric

    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.


    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.

    .tensorflow.metadata.v0.PerformanceMetric performance_metric = 3;

    returns

    The performanceMetric.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  12. def getPerformanceMetricOrBuilder(): PerformanceMetricOrBuilder

    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.


    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.

    .tensorflow.metadata.v0.PerformanceMetric performance_metric = 3;

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  13. def getRepeatedField(field: FieldDescriptor, index: Int): AnyRef
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  14. def getRepeatedFieldCount(field: FieldDescriptor): Int
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  15. def getSerializedSize(): Int
    Definition Classes
    MetaOptimizationTarget → GeneratedMessageV3 → AbstractMessage → MessageLite
    Annotations
    @Override()
  16. def getTaskName(): String

    The name of a task in this problem statement producing the
prediction or classification for the metric.


    The name of a task in this problem statement producing the
prediction or classification for the metric.

    string task_name = 1;

    returns

    The taskName.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  17. def getTaskNameBytes(): ByteString

    The name of a task in this problem statement producing the
prediction or classification for the metric.


    The name of a task in this problem statement producing the
prediction or classification for the metric.

    string task_name = 1;

    returns

    The bytes for taskName.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  18. def getThresholdConfig(): ThresholdConfig

    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.


    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.

    .tensorflow.metadata.v0.MetaOptimizationTarget.ThresholdConfig threshold_config = 5;

    returns

    The thresholdConfig.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  19. def getThresholdConfigOrBuilder(): ThresholdConfigOrBuilder

    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.


    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.

    .tensorflow.metadata.v0.MetaOptimizationTarget.ThresholdConfig threshold_config = 5;

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  20. def getUnknownFields(): UnknownFieldSet
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  21. def hasField(field: FieldDescriptor): Boolean
    Definition Classes
    GeneratedMessageV3 → MessageOrBuilder
  22. def hasOneof(oneof: OneofDescriptor): Boolean
    Definition Classes
    GeneratedMessageV3 → AbstractMessage → MessageOrBuilder
  23. def hasPerformanceMetric(): Boolean

    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.


    The performance metric to be evaluated.
The prediction or classification is based upon the task.
The label is from the type of the task, or from the override_task.

    .tensorflow.metadata.v0.PerformanceMetric performance_metric = 3;

    returns

    Whether the performanceMetric field is set.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  24. def hasThresholdConfig(): Boolean

    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.


    Secondary meta optimization targets can be thresholded, meaning that the
optimization process prefers solutions above (or below) the threshold,
but need not prefer solutions higher (or lower) on the metric if the
threshold is met.

    .tensorflow.metadata.v0.MetaOptimizationTarget.ThresholdConfig threshold_config = 5;

    returns

    Whether the thresholdConfig field is set.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override()
  25. def hashCode(): Int
    Definition Classes
    MetaOptimizationTarget → AbstractMessage → Message → AnyRef → Any
    Annotations
    @Override()
  26. final def isInitialized(): Boolean
    Definition Classes
    MetaOptimizationTarget → GeneratedMessageV3 → AbstractMessage → MessageLiteOrBuilder
    Annotations
    @Override()
  27. def newBuilderForType(): Builder
    Definition Classes
    MetaOptimizationTarget → Message → MessageLite
    Annotations
    @Override()
  28. def toBuilder(): Builder
    Definition Classes
    MetaOptimizationTarget → Message → MessageLite
    Annotations
    @Override()
  29. def toByteArray(): Array[Byte]
    Definition Classes
    AbstractMessageLite → MessageLite
  30. def toByteString(): ByteString
    Definition Classes
    AbstractMessageLite → MessageLite
  31. final def toString(): String
    Definition Classes
    AbstractMessage → Message → AnyRef → Any
  32. def writeDelimitedTo(output: OutputStream): Unit
    Definition Classes
    AbstractMessageLite → MessageLite
    Annotations
    @throws(classOf[java.io.IOException])
  33. def writeTo(output: CodedOutputStream): Unit
    Definition Classes
    MetaOptimizationTarget → GeneratedMessageV3 → AbstractMessage → MessageLite
    Annotations
    @Override()
  34. def writeTo(output: OutputStream): Unit
    Definition Classes
    AbstractMessageLite → MessageLite
    Annotations
    @throws(classOf[java.io.IOException])

Deprecated Value Members

  1. def getWeight(): Double

    If a model spec has multiple meta optimization targets, the weight
of each can be specified. The final objective is then a weighted
combination of the multiple objectives. If not specified, value is 1.


    If a model spec has multiple meta optimization targets, the weight
of each can be specified. The final objective is then a weighted
combination of the multiple objectives. If not specified, value is 1.

    double weight = 4 [deprecated = true];

    returns

    The weight.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override() @Deprecated
    Deprecated
  2. def hasWeight(): Boolean

    If a model spec has multiple meta optimization targets, the weight
of each can be specified. The final objective is then a weighted
combination of the multiple objectives. If not specified, value is 1.


    If a model spec has multiple meta optimization targets, the weight
of each can be specified. The final objective is then a weighted
combination of the multiple objectives. If not specified, value is 1.

    double weight = 4 [deprecated = true];

    returns

    Whether the weight field is set.

    Definition Classes
    MetaOptimizationTarget → MetaOptimizationTargetOrBuilder
    Annotations
    @Override() @Deprecated
    Deprecated