VariableAggregation
Stay organized with collections
Save and categorize content based on your preferences.
Indicates how a distributed variable will be aggregated.
Protobuf enum
tensorflow.VariableAggregation
Inherited Methods
From class
java.lang.Enum
final
int
|
|
int
|
compareTo(Object arg0)
|
final
boolean
|
equals(Object arg0)
|
final
Class<VariableAggregation>
|
getDeclaringClass()
|
final
int
|
hashCode()
|
final
String
|
name()
|
final
int
|
ordinal()
|
String
|
toString()
|
static
<T extends Enum<T>>
T
|
valueOf(Class<T> arg0, String arg1)
|
From class
java.lang.Object
boolean
|
equals(Object arg0)
|
final
Class<?>
|
getClass()
|
int
|
hashCode()
|
final
void
|
notify()
|
final
void
|
notifyAll()
|
String
|
toString()
|
final
void
|
wait(long arg0, int arg1)
|
final
void
|
wait(long arg0)
|
final
void
|
wait()
|
From interface
java.lang.Comparable
Enum Values
public
static
final
VariableAggregation
UNRECOGNIZED
public
static
final
VariableAggregation
VARIABLE_AGGREGATION_MEAN
`MEAN`: Take the arithmetic mean ("average") of the updates across
replicas.
VARIABLE_AGGREGATION_MEAN = 2;
public
static
final
VariableAggregation
VARIABLE_AGGREGATION_NONE
`NONE`: This is the default, giving an error if you use a
variable-update operation with multiple replicas.
VARIABLE_AGGREGATION_NONE = 0;
public
static
final
VariableAggregation
VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA
`ONLY_FIRST_REPLICA`: This is for when every replica is performing the same
update, but we only want to perform the update once. Used, e.g., for the
global step counter.
VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA = 3;
public
static
final
VariableAggregation
VARIABLE_AGGREGATION_SUM
`SUM`: Add the updates across replicas.
VARIABLE_AGGREGATION_SUM = 1;
Constants
public
static
final
int
VARIABLE_AGGREGATION_MEAN_VALUE
`MEAN`: Take the arithmetic mean ("average") of the updates across
replicas.
VARIABLE_AGGREGATION_MEAN = 2;
Constant Value:
2
public
static
final
int
VARIABLE_AGGREGATION_NONE_VALUE
`NONE`: This is the default, giving an error if you use a
variable-update operation with multiple replicas.
VARIABLE_AGGREGATION_NONE = 0;
Constant Value:
0
public
static
final
int
VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA_VALUE
`ONLY_FIRST_REPLICA`: This is for when every replica is performing the same
update, but we only want to perform the update once. Used, e.g., for the
global step counter.
VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA = 3;
Constant Value:
3
public
static
final
int
VARIABLE_AGGREGATION_SUM_VALUE
`SUM`: Add the updates across replicas.
VARIABLE_AGGREGATION_SUM = 1;
Constant Value:
1
Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2021-11-29 UTC.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2021-11-29 UTC."],[],[],null,["# VariableAggregation\n\npublic final enum **VariableAggregation** \n\u003cbr /\u003e\n\n```\n Indicates how a distributed variable will be aggregated.\n \n```\nProtobuf enum `tensorflow.VariableAggregation`\n\n\u003cbr /\u003e\n\n### Constants\n\n|-----|----------------------------------------------------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------|\n| int | [VARIABLE_AGGREGATION_MEAN_VALUE](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation#VARIABLE_AGGREGATION_MEAN_VALUE) | ``` `MEAN`: Take the arithmetic mean (\"average\") of the updates across replicas. ``` |\n| int | [VARIABLE_AGGREGATION_NONE_VALUE](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation#VARIABLE_AGGREGATION_NONE_VALUE) | ``` `NONE`: This is the default, giving an error if you use a variable-update operation with multiple replicas. ``` |\n| int | [VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA_VALUE](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation#VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA_VALUE) | ``` `ONLY_FIRST_REPLICA`: This is for when every replica is performing the same update, but we only want to perform the update once. ``` |\n| int | [VARIABLE_AGGREGATION_SUM_VALUE](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation#VARIABLE_AGGREGATION_SUM_VALUE) | ``` `SUM`: Add the updates across replicas. ``` |\n\n### Inherited Methods\n\nFrom class java.lang.Enum \n\n|-------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------|\n| final int | compareTo([VariableAggregation](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation) arg0) |\n| int | compareTo(Object arg0) |\n| final boolean | equals(Object arg0) |\n| final Class\\\u003c[VariableAggregation](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation)\\\u003e | getDeclaringClass() |\n| final int | hashCode() |\n| final String | name() |\n| final int | ordinal() |\n| String | toString() |\n| static \\\u003cT extends Enum\\\u003cT\\\u003e\\\u003e T | valueOf(Class\\\u003cT\\\u003e arg0, String arg1) |\n\nFrom class java.lang.Object \n\n|------------------|---------------------------|\n| boolean | equals(Object arg0) |\n| final Class\\\u003c?\\\u003e | getClass() |\n| int | hashCode() |\n| final void | notify() |\n| final void | notifyAll() |\n| String | toString() |\n| final void | wait(long arg0, int arg1) |\n| final void | wait(long arg0) |\n| final void | wait() |\n\nFrom interface java.lang.Comparable \n\n|--------------|--------------------------------------------------------------------------------------------------------------|\n| abstract int | compareTo([VariableAggregation](/jvm/api_docs/java/org/tensorflow/proto/framework/VariableAggregation) arg0) |\n\nEnum Values\n-----------\n\n#### public static final VariableAggregation\n**UNRECOGNIZED**\n\n\u003cbr /\u003e\n\n#### public static final VariableAggregation\n**VARIABLE_AGGREGATION_MEAN**\n\n\u003cbr /\u003e\n\n```\n `MEAN`: Take the arithmetic mean (\"average\") of the updates across\n replicas.\n \n```\n`VARIABLE_AGGREGATION_MEAN = 2;`\n\n\u003cbr /\u003e\n\n#### public static final VariableAggregation\n**VARIABLE_AGGREGATION_NONE**\n\n\u003cbr /\u003e\n\n```\n `NONE`: This is the default, giving an error if you use a\n variable-update operation with multiple replicas.\n \n```\n`VARIABLE_AGGREGATION_NONE = 0;`\n\n\u003cbr /\u003e\n\n#### public static final VariableAggregation\n**VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA**\n\n\u003cbr /\u003e\n\n```\n `ONLY_FIRST_REPLICA`: This is for when every replica is performing the same\n update, but we only want to perform the update once. Used, e.g., for the\n global step counter.\n \n```\n`VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA = 3;`\n\n\u003cbr /\u003e\n\n#### public static final VariableAggregation\n**VARIABLE_AGGREGATION_SUM**\n\n\u003cbr /\u003e\n\n```\n `SUM`: Add the updates across replicas.\n \n```\n`VARIABLE_AGGREGATION_SUM = 1;`\n\n\u003cbr /\u003e\n\nConstants\n---------\n\n#### public static final int\n**VARIABLE_AGGREGATION_MEAN_VALUE**\n\n\u003cbr /\u003e\n\n```\n `MEAN`: Take the arithmetic mean (\"average\") of the updates across\n replicas.\n \n```\n`VARIABLE_AGGREGATION_MEAN = 2;`\n\n\u003cbr /\u003e\n\nConstant Value: 2 \n\n#### public static final int\n**VARIABLE_AGGREGATION_NONE_VALUE**\n\n\u003cbr /\u003e\n\n```\n `NONE`: This is the default, giving an error if you use a\n variable-update operation with multiple replicas.\n \n```\n`VARIABLE_AGGREGATION_NONE = 0;`\n\n\u003cbr /\u003e\n\nConstant Value: 0 \n\n#### public static final int\n**VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA_VALUE**\n\n\u003cbr /\u003e\n\n```\n `ONLY_FIRST_REPLICA`: This is for when every replica is performing the same\n update, but we only want to perform the update once. Used, e.g., for the\n global step counter.\n \n```\n`VARIABLE_AGGREGATION_ONLY_FIRST_REPLICA = 3;`\n\n\u003cbr /\u003e\n\nConstant Value: 3 \n\n#### public static final int\n**VARIABLE_AGGREGATION_SUM_VALUE**\n\n\u003cbr /\u003e\n\n```\n `SUM`: Add the updates across replicas.\n \n```\n`VARIABLE_AGGREGATION_SUM = 1;`\n\n\u003cbr /\u003e\n\nConstant Value: 1"]]