llvm::Logger Class Reference

Logging utility - given an ordered specification of features, and assuming a scalar reward, allow logging feature values and rewards. More...

#include "llvm/Analysis/Utils/TrainingLogger.h"

Public Member Functions
LLVM_ABI	Logger (std::unique_ptr< raw_ostream > OS, const std::vector< TensorSpec > &FeatureSpecs, const TensorSpec &RewardSpec, bool IncludeReward, std::optional< TensorSpec > AdviceSpec=std::nullopt)
	Construct a Logger.
LLVM_ABI void	switchContext (StringRef Name)
LLVM_ABI void	startObservation ()
LLVM_ABI void	endObservation ()
void	flush ()
const std::string &	currentContext () const
bool	hasObservationInProgress () const
	Check if there is at least an observation for currentContext().
bool	hasAnyObservationForContext (StringRef Ctx) const
	Check if there is at least an observation for the context Ctx.
template<typename T>
void	logReward (T Value)
void	logTensorValue (size_t FeatureID, const char *RawData)

Detailed Description

Logging utility - given an ordered specification of features, and assuming a scalar reward, allow logging feature values and rewards.

The assumption is that, for an event to be logged (i.e. a set of feature values and a reward), the user calls the log* API for each feature exactly once, providing the index matching the position in the feature spec list provided at construction. The example assumes the first feature's element type is float, the second is int64, and the reward is float:

event 0: logFloatValue(0, ...) logInt64Value(1, ...) ... logFloatReward(...) event 1: logFloatValue(0, ...) logInt64Value(1, ...) ... logFloatReward(...)

At the end, call print to generate the log. Alternatively, don't call logReward at the end of each event, just log{Float|Int32|Int64}FinalReward at the end.

Definition at line 92 of file TrainingLogger.h.

Constructor & Destructor Documentation

Logger()

Logger::Logger	(	std::unique_ptr< raw_ostream >	OS,
		const std::vector< TensorSpec > &	FeatureSpecs,
		const TensorSpec &	RewardSpec,
		bool	IncludeReward,
		std::optional< TensorSpec >	AdviceSpec = std::nullopt )

Construct a Logger.

If IncludeReward is false, then logReward or logFinalReward shouldn't be called, and the reward feature won't be printed out. NOTE: the FeatureSpecs are expected to be in the same order (i.e. have corresponding indices) with any MLModelRunner implementations corresponding to the model being trained/logged.

Definition at line 82 of file TrainingLogger.cpp.

References llvm::move().

Member Function Documentation

currentContext()

const std::string & llvm::Logger::currentContext ( ) const

inline

Definition at line 123 of file TrainingLogger.h.

endObservation()

void Logger::endObservation

(

)

Definition at line 68 of file TrainingLogger.cpp.

flush()

void llvm::Logger::flush ( )

inline

Definition at line 121 of file TrainingLogger.h.

hasAnyObservationForContext()

bool llvm::Logger::hasAnyObservationForContext ( StringRef Ctx ) const

inline

Check if there is at least an observation for the context Ctx.

Definition at line 131 of file TrainingLogger.h.

Referenced by hasObservationInProgress().

hasObservationInProgress()

bool llvm::Logger::hasObservationInProgress ( ) const

inline

Check if there is at least an observation for currentContext().

Definition at line 126 of file TrainingLogger.h.

References hasAnyObservationForContext().

logReward()

template<typename T>

void llvm::Logger::logReward ( T Value )

inline

Definition at line 135 of file TrainingLogger.h.

References T.

logTensorValue()

void llvm::Logger::logTensorValue ( size_t FeatureID, const char * RawData )

inline

Definition at line 139 of file TrainingLogger.h.

startObservation()

void Logger::startObservation

(

)

Definition at line 58 of file TrainingLogger.cpp.

References llvm::json::OStream::attribute(), I, and llvm::json::OStream::object().

switchContext()

void Logger::switchContext

(

StringRef

Name

)

Definition at line 51 of file TrainingLogger.cpp.

References llvm::json::OStream::attribute(), and llvm::json::OStream::object().

---

The documentation for this class was generated from the following files:

• include/llvm/Analysis/Utils/TrainingLogger.h
• lib/Analysis/TrainingLogger.cpp