Logo ROOT  
Reference Guide
TMVA::DNN::TDLGradientDescent< Architecture_t > Class Template Reference

template<typename Architecture_t>
class TMVA::DNN::TDLGradientDescent< Architecture_t >

Definition at line 65 of file DLMinimizers.h.

Public Types

using DeepNet_t = TDeepNet< Architecture_t >
 
using Matrix_t = typename Architecture_t::Matrix_t
 
using Scalar_t = typename Architecture_t::Scalar_t
 

Public Member Functions

 TDLGradientDescent ()
 
 TDLGradientDescent (Scalar_t learningRate, size_t convergenceSteps, size_t testInterval)
 
size_t GetConvergenceCount () const
 Getters. More...
 
size_t GetConvergenceSteps () const
 
Scalar_t GetTestError () const
 
size_t GetTestInterval () const
 
Scalar_t GetTrainingError () const
 
bool HasConverged ()
 Increases the minimization step counter by the test error evaluation period and uses the current internal value of the test error to determine if the minimization has converged. More...
 
bool HasConverged (Scalar_t testError)
 Increases the minimization step counter by the test error evaluation period and uses the provided test error value to determine if the minimization has converged. More...
 
void Reset ()
 Reset minimizer object to default state. More...
 
void SetBatchSize (Scalar_t rate)
 
void SetConvergenceSteps (size_t steps)
 Setters. More...
 
void SetLearningRate (Scalar_t rate)
 
void SetTestInterval (size_t interval)
 
void Step (DeepNet_t &deepNet, std::vector< Matrix_t > &input, const Matrix_t &output, const Matrix_t &weights)
 Perform a single optimization step on a given batch. More...
 
void Step (DeepNet_t &master, std::vector< DeepNet_t > &nets, std::vector< TTensorBatch< Architecture_t > > &batches)
 Perform multiple optimization steps simultaneously. More...
 
Scalar_t StepLoss (DeepNet_t &deepNet, std::vector< Matrix_t > &input, const Matrix_t &output, const Matrix_t &weights)
 Same as Step(...) but also evaluate the loss on the given training data. More...
 
void StepMomentum (DeepNet_t &master, std::vector< DeepNet_t > &nets, std::vector< TTensorBatch< Architecture_t > > &batches, Scalar_t momentum)
 Same as the Step(...) method for multiple batches but uses momentum. More...
 
void StepNesterov (DeepNet_t &master, std::vector< DeepNet_t > &nets, std::vector< TTensorBatch< Architecture_t > > &batches, Scalar_t momentum)
 Same as the Step(...) method for multiple batches but uses Nesterov momentum. More...
 
void StepReducedWeights (DeepNet_t &deepNet, std::vector< Matrix_t > &input, const Matrix_t &output, const Matrix_t &weights)
 Does not evaluate the loss and therefore not trigger a possible synchronization with the device. More...
 
Scalar_t StepReducedWeightsLoss (DeepNet_t &deepNet, std::vector< Matrix_t > &input, const Matrix_t &output, const Matrix_t &weights)
 Similar to StepReducedWeights(...) but also evaluates the loss. More...
 

Private Attributes

size_t fBatchSize
 Batch size to use for the training. More...
 
size_t fConvergenceCount
 Current number of training epochs without. More...
 
size_t fConvergenceSteps
 Number of training epochs without considerable. More...
 
Scalar_t fLearningRate
 Learning rate \(\alpha\). More...
 
Scalar_t fMinimumError
 The minimum loss achieved on the training set. More...
 
size_t fStepCount
 Number of steps performed in the current training session. More...
 
Scalar_t fTestError
 Holds the most recently computed test loss. More...
 
size_t fTestInterval
 Interval for the computation of the test error. More...
 
Scalar_t fTrainingError
 Holds the most recently computed training loss. More...
 

#include <TMVA/DNN/DLMinimizers.h>

Member Typedef Documentation

◆ DeepNet_t

template<typename Architecture_t >
using TMVA::DNN::TDLGradientDescent< Architecture_t >::DeepNet_t = TDeepNet<Architecture_t>

Definition at line 67 of file DLMinimizers.h.

◆ Matrix_t

template<typename Architecture_t >
using TMVA::DNN::TDLGradientDescent< Architecture_t >::Matrix_t = typename Architecture_t::Matrix_t

Definition at line 69 of file DLMinimizers.h.

◆ Scalar_t

template<typename Architecture_t >
using TMVA::DNN::TDLGradientDescent< Architecture_t >::Scalar_t = typename Architecture_t::Scalar_t

Definition at line 68 of file DLMinimizers.h.

Constructor & Destructor Documentation

◆ TDLGradientDescent() [1/2]

template<typename Architecture_t >
TMVA::DNN::TDLGradientDescent< Architecture_t >::TDLGradientDescent

Definition at line 164 of file DLMinimizers.h.

◆ TDLGradientDescent() [2/2]

template<typename Architecture_t >
TMVA::DNN::TDLGradientDescent< Architecture_t >::TDLGradientDescent ( Scalar_t  learningRate,
size_t  convergenceSteps,
size_t  testInterval 
)

Definition at line 173 of file DLMinimizers.h.

Member Function Documentation

◆ GetConvergenceCount()

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::GetConvergenceCount ( ) const
inline

Getters.

Definition at line 147 of file DLMinimizers.h.

◆ GetConvergenceSteps()

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::GetConvergenceSteps ( ) const
inline

Definition at line 148 of file DLMinimizers.h.

◆ GetTestError()

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::GetTestError ( ) const
inline

Definition at line 150 of file DLMinimizers.h.

◆ GetTestInterval()

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::GetTestInterval ( ) const
inline

Definition at line 151 of file DLMinimizers.h.

◆ GetTrainingError()

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::GetTrainingError ( ) const
inline

Definition at line 149 of file DLMinimizers.h.

◆ HasConverged() [1/2]

template<typename Architecture_t >
bool TMVA::DNN::TDLGradientDescent< Architecture_t >::HasConverged

Increases the minimization step counter by the test error evaluation period and uses the current internal value of the test error to determine if the minimization has converged.

Definition at line 277 of file DLMinimizers.h.

◆ HasConverged() [2/2]

template<typename Architecture_t >
bool TMVA::DNN::TDLGradientDescent< Architecture_t >::HasConverged ( Scalar_t  testError)

Increases the minimization step counter by the test error evaluation period and uses the provided test error value to determine if the minimization has converged.

Definition at line 291 of file DLMinimizers.h.

◆ Reset()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::Reset ( void  )
inline

Reset minimizer object to default state.

Definition at line 90 of file DLMinimizers.h.

◆ SetBatchSize()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::SetBatchSize ( Scalar_t  rate)
inline

Definition at line 157 of file DLMinimizers.h.

◆ SetConvergenceSteps()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::SetConvergenceSteps ( size_t  steps)
inline

Setters.

Definition at line 154 of file DLMinimizers.h.

◆ SetLearningRate()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::SetLearningRate ( Scalar_t  rate)
inline

Definition at line 156 of file DLMinimizers.h.

◆ SetTestInterval()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::SetTestInterval ( size_t  interval)
inline

Definition at line 155 of file DLMinimizers.h.

◆ Step() [1/2]

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::Step ( DeepNet_t deepNet,
std::vector< Matrix_t > &  input,
const Matrix_t output,
const Matrix_t weights 
)

Perform a single optimization step on a given batch.

Propagates the input matrix foward through the net, evaluates the loss and propagates the gradients backward through the net. The computed gradients are scaled by the learning rate \(\alpha\) and subtracted from the weights and bias values of each layer.

Definition at line 183 of file DLMinimizers.h.

◆ Step() [2/2]

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::Step ( DeepNet_t master,
std::vector< DeepNet_t > &  nets,
std::vector< TTensorBatch< Architecture_t > > &  batches 
)

Perform multiple optimization steps simultaneously.

Performs the backprop algorithm on the input batches given in batches on the neural networks given in nets. The forward and backward propagation steps are executed in an interleaving manner in order to exploit potential batch-level parallelism for asynchronous device calls.

Definition at line 247 of file DLMinimizers.h.

◆ StepLoss()

template<typename Architecture_t >
auto TMVA::DNN::TDLGradientDescent< Architecture_t >::StepLoss ( DeepNet_t deepNet,
std::vector< Matrix_t > &  input,
const Matrix_t output,
const Matrix_t weights 
)

Same as Step(...) but also evaluate the loss on the given training data.

Note that this requires synchronization between host and device.

Definition at line 213 of file DLMinimizers.h.

◆ StepMomentum()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::StepMomentum ( DeepNet_t master,
std::vector< DeepNet_t > &  nets,
std::vector< TTensorBatch< Architecture_t > > &  batches,
Scalar_t  momentum 
)

Same as the Step(...) method for multiple batches but uses momentum.

Definition at line 257 of file DLMinimizers.h.

◆ StepNesterov()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::StepNesterov ( DeepNet_t master,
std::vector< DeepNet_t > &  nets,
std::vector< TTensorBatch< Architecture_t > > &  batches,
Scalar_t  momentum 
)

Same as the Step(...) method for multiple batches but uses Nesterov momentum.

Definition at line 267 of file DLMinimizers.h.

◆ StepReducedWeights()

template<typename Architecture_t >
void TMVA::DNN::TDLGradientDescent< Architecture_t >::StepReducedWeights ( DeepNet_t deepNet,
std::vector< Matrix_t > &  input,
const Matrix_t output,
const Matrix_t weights 
)

Does not evaluate the loss and therefore not trigger a possible synchronization with the device.

Trains the weights of each layer, but only the bias terms of the first layer for compatibility with the previous implementation.

Definition at line 194 of file DLMinimizers.h.

◆ StepReducedWeightsLoss()

template<typename Architecture_t >
auto TMVA::DNN::TDLGradientDescent< Architecture_t >::StepReducedWeightsLoss ( DeepNet_t deepNet,
std::vector< Matrix_t > &  input,
const Matrix_t output,
const Matrix_t weights 
)

Similar to StepReducedWeights(...) but also evaluates the loss.

May trigger synchronization with the device.

Definition at line 225 of file DLMinimizers.h.

Member Data Documentation

◆ fBatchSize

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fBatchSize
private

Batch size to use for the training.

Definition at line 72 of file DLMinimizers.h.

◆ fConvergenceCount

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fConvergenceCount
private

Current number of training epochs without.

considerable decrease in the test error.

Definition at line 76 of file DLMinimizers.h.

◆ fConvergenceSteps

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fConvergenceSteps
private

Number of training epochs without considerable.

decrease in the test error for convergence.

Definition at line 74 of file DLMinimizers.h.

◆ fLearningRate

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fLearningRate
private

Learning rate \(\alpha\).

Definition at line 81 of file DLMinimizers.h.

◆ fMinimumError

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fMinimumError
private

The minimum loss achieved on the training set.

during the current traning session.

Definition at line 82 of file DLMinimizers.h.

◆ fStepCount

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fStepCount
private

Number of steps performed in the current training session.

Definition at line 73 of file DLMinimizers.h.

◆ fTestError

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fTestError
private

Holds the most recently computed test loss.

Definition at line 80 of file DLMinimizers.h.

◆ fTestInterval

template<typename Architecture_t >
size_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fTestInterval
private

Interval for the computation of the test error.

Definition at line 78 of file DLMinimizers.h.

◆ fTrainingError

template<typename Architecture_t >
Scalar_t TMVA::DNN::TDLGradientDescent< Architecture_t >::fTrainingError
private

Holds the most recently computed training loss.

Definition at line 79 of file DLMinimizers.h.


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