EventAccumulator

Accumulates event values collected from the provided path.

The EventAccumulator is intended to provide a convenient interface for loading event data written during a TensorFlow run (or otherwise). TensorFlow writes out event ProtoBuf objects, which have a timestamp and step number associated with them, and often also contain a Summary. Summaries can store different kinds of data like a scalar value, an image, audio, or a histogram. Each summary also has a tag associated with it, which we use to organize logically related data. The EventAccumulator supports retrieving the event and summary data by their tags.

Calling tags returns a map from event types to the associated tags for those types, that were found in the loaded event files. Then, various functional endpoints (e.g., scalars(tag)) allow for the retrieval of all data associated with each tag.

The reload() method synchronously loads all of the data written so far.

path: Path to a directory containing TensorFlow events files, or a single TensorFlow events file. The accumulator will load events from this path.
sizeGuidance: Information on how much data the event accumulator should store in memory. The default size guidance tries not to store too much so as to avoid consuming all of the client's memory. The sizeGuidance should be a map from event types to integers representing the number of items to keep in memory, per tag for items of that event type. If the size is 0, then all events are stored. Images, audio, and histograms tend to be very large and thus storing all of them is not recommended.
histogramCompressionBps: Information on how the event accumulator should compress histogram data for the CompressedHistogramEventType event type.
purgeOrphanedData: Boolean value indicating whether to discard any events that were "orphaned" by a TensorFlow restart.

Linear Supertypes

Serializable, Serializable, Product, Equals, AnyRef, Any

Instance Constructors

new EventAccumulator(path: Path, sizeGuidance: Map[EventType, Int] = ..., histogramCompressionBps: Seq[Int] = ..., purgeOrphanedData: Boolean = true)

path
Path to a directory containing TensorFlow events files, or a single TensorFlow events file. The accumulator will load events from this path.
sizeGuidance
Information on how much data the event accumulator should store in memory. The default size guidance tries not to store too much so as to avoid consuming all of the client's memory. The sizeGuidance should be a map from event types to integers representing the number of items to keep in memory, per tag for items of that event type. If the size is 0, then all events are stored. Images, audio, and histograms tend to be very large and thus storing all of them is not recommended.
histogramCompressionBps
Information on how the event accumulator should compress histogram data for the CompressedHistogramEventType event type.
purgeOrphanedData
Boolean value indicating whether to discard any events that were "orphaned" by a TensorFlow restart.

Value Members

final def !=(arg0: Any): Boolean

Definition Classes
AnyRef → Any
final def ##(): Int

Definition Classes
AnyRef → Any
final def ==(arg0: Any): Boolean

Definition Classes
AnyRef → Any
final def asInstanceOf[T0]: T0

Definition Classes
Any
def audio(tag: String): List[AudioEventRecord]

Returns all audio events associated with the provided summary tag.
def clone(): AnyRef

Attributes
protected[java.lang]
Definition Classes
AnyRef
Annotations
@throws( ... )
def compressedHistograms(tag: String): List[CompressedHistogramEventRecord]

Returns all compressed histogram events associated with the provided summary tag.
final def eq(arg0: AnyRef): Boolean

Definition Classes
AnyRef
def finalize(): Unit

Attributes
protected[java.lang]
Definition Classes
AnyRef
Annotations
@throws( classOf[java.lang.Throwable] )
def firstEventTimeStamp: Double

Returns the timestamp (in seconds) of the first event.
Returns the timestamp (in seconds) of the first event.
If the first event has been loaded (either by this method or by reload(), then this method returns immediately. Otherwise, it loads the first event and then returns. Note that this means that calling reload() will cause this method to block until reload() has finished.
final def getClass(): Class[_]

Definition Classes
AnyRef → Any
def graph: GraphDef

Returns the graph definition, if there is one.
Returns the graph definition, if there is one.
If the graph is stored directly, the method returns it. If no graph is stored directly, but a meta-graph is stored containing a graph, the method returns that graph.

Annotations
@throws( ... )
val histogramCompressionBps: Seq[Int]

Information on how the event accumulator should compress histogram data for the CompressedHistogramEventType event type.
def histograms(tag: String): List[HistogramEventRecord]

Returns all histogram events associated with the provided summary tag.
def images(tag: String): List[ImageEventRecord]

Returns all image events associated with the provided summary tag.
final def isInstanceOf[T0]: Boolean

Definition Classes
Any
def metaGraph: MetaGraphDef

Returns the meta-graph definition, if there is one.
Returns the meta-graph definition, if there is one.

Annotations
@throws( ... )
final def ne(arg0: AnyRef): Boolean

Definition Classes
AnyRef
final def notify(): Unit

Definition Classes
AnyRef
final def notifyAll(): Unit

Definition Classes
AnyRef
val path: Path

Path to a directory containing TensorFlow events files, or a single TensorFlow events file.
Path to a directory containing TensorFlow events files, or a single TensorFlow events file. The accumulator will load events from this path.
def pluginAssets(pluginName: String): Seq[Path]

Returns a sequence with paths to all the registered assets for the provided plugin name.
Returns a sequence with paths to all the registered assets for the provided plugin name.
If a plugins directory does not exist in the managed directory, then this method returns an empty list. This maintains compatibility with old log directories that contain no plugin sub-directories.
def pluginTagToContent(pluginName: String): Option[Map[String, String]]

Returns a map from tags to content specific to the specified plugin.
val purgeOrphanedData: Boolean

Boolean value indicating whether to discard any events that were "orphaned" by a TensorFlow restart.
def reload(): EventAccumulator

Loads all events added since the last call to reload() and returns this event accumulator.
Loads all events added since the last call to reload() and returns this event accumulator. If reload() was never called before, then it loads all events in the path.
def retrievePluginAsset(pluginName: String, assetName: String): String

Retrieves a particular plugin asset from the managed directory and returns it as a string.
def runMetadata(tag: String): RunMetadata

Returns the run metadata associated with the provided summary tag.
Returns the run metadata associated with the provided summary tag.

Annotations
@throws( ... )
def scalars(tag: String): List[ScalarEventRecord]

Returns all scalar events associated with the provided summary tag.
val sizeGuidance: Map[EventType, Int]

Information on how much data the event accumulator should store in memory.
Information on how much data the event accumulator should store in memory. The default size guidance tries not to store too much so as to avoid consuming all of the client's memory. The sizeGuidance should be a map from event types to integers representing the number of items to keep in memory, per tag for items of that event type. If the size is 0, then all events are stored. Images, audio, and histograms tend to be very large and thus storing all of them is not recommended.
def summaryMetadata(tag: String): SummaryMetadata

Returns the summary metadata associated with the provided summary tag.
final def synchronized[T0](arg0: ⇒ T0): T0

Definition Classes
AnyRef
def tags: Map[EventType, Seq[String]]

Returns a map from event types to all corresponding tags that have been accumulated.
def tensors(tag: String): List[TensorEventRecord]

Returns all tensor events associated with the provided summary tag.
final def wait(): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )
final def wait(arg0: Long, arg1: Int): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )
final def wait(arg0: Long): Unit

Definition Classes
AnyRef
Annotations
@throws( ... )

Related Docs: object EventAccumulator | package events

case class EventAccumulator(path: Path, sizeGuidance: Map[EventType, Int] = ..., histogramCompressionBps: Seq[Int] = ..., purgeOrphanedData: Boolean = true) extends Product with Serializable

Instance Constructors

new EventAccumulator(path: Path, sizeGuidance: Map[EventType, Int] = ..., histogramCompressionBps: Seq[Int] = ..., purgeOrphanedData: Boolean = true)

Value Members

final def !=(arg0: Any): Boolean

final def ##(): Int

final def ==(arg0: Any): Boolean

final def asInstanceOf[T0]: T0

def audio(tag: String): List[AudioEventRecord]

def clone(): AnyRef

def compressedHistograms(tag: String): List[CompressedHistogramEventRecord]

final def eq(arg0: AnyRef): Boolean

def finalize(): Unit

def firstEventTimeStamp: Double

final def getClass(): Class[_]

def graph: GraphDef

val histogramCompressionBps: Seq[Int]

def histograms(tag: String): List[HistogramEventRecord]

def images(tag: String): List[ImageEventRecord]

final def isInstanceOf[T0]: Boolean

def metaGraph: MetaGraphDef

final def ne(arg0: AnyRef): Boolean

final def notify(): Unit

final def notifyAll(): Unit

val path: Path

def pluginAssets(pluginName: String): Seq[Path]

def pluginTagToContent(pluginName: String): Option[Map[String, String]]

val purgeOrphanedData: Boolean

def reload(): EventAccumulator

def retrievePluginAsset(pluginName: String, assetName: String): String

def runMetadata(tag: String): RunMetadata

def scalars(tag: String): List[ScalarEventRecord]

val sizeGuidance: Map[EventType, Int]

def summaryMetadata(tag: String): SummaryMetadata

final def synchronized[T0](arg0: ⇒ T0): T0

def tags: Map[EventType, Seq[String]]

def tensors(tag: String): List[TensorEventRecord]

final def wait(): Unit

final def wait(arg0: Long, arg1: Int): Unit

final def wait(arg0: Long): Unit

Inherited from Serializable

Inherited from Serializable

Inherited from Product

Inherited from Equals

Inherited from AnyRef

Inherited from Any

Ungrouped