mxnet
Public Member Functions | List of all members
mxnet::cpp::AdaGradOptimizer Class Reference

#include <optimizer.h>

Inheritance diagram for mxnet::cpp::AdaGradOptimizer:
Inheritance graph
Collaboration diagram for mxnet::cpp::AdaGradOptimizer:
Collaboration graph

Public Member Functions

 AdaGradOptimizer (unsigned begin_num_update=0)
 
std::string GetType () const override
 get optimizer type More...
 
void Update (int index, NDArray weight, NDArray grad) override
 Update a weight with gradient. More...
 
- Public Member Functions inherited from mxnet::cpp::Optimizer
 Optimizer (unsigned begin_num_update)
 constructor More...
 
virtual ~Optimizer ()
 destructor More...
 
template<typename T >
OptimizerSetParam (const std::string &name, const T &value)
 set config parameters More...
 
OptimizerSetLRScheduler (std::unique_ptr< LRScheduler > lrScheduler)
 
std::string Serialize () const
 Serialize the optimizer parameters to a string. More...
 

Additional Inherited Members

- Protected Member Functions inherited from mxnet::cpp::Optimizer
const std::vector< const char * > GetParamKeys_ () const
 
const std::vector< const char * > GetParamValues_ () const
 
unsigned UpdateCount_ (int index)
 
float GetLR_ (int index)
 
float GetWD_ (int index)
 
- Static Protected Member Functions inherited from mxnet::cpp::Optimizer
static OpMap *& op_map ()
 
- Protected Attributes inherited from mxnet::cpp::Optimizer
std::map< std::string, std::string > params_
 
std::map< int, unsigned > count_
 
unsigned begin_num_update_
 
unsigned num_update_
 
std::unique_ptr< LRSchedulerlrScheduler_ = nullptr
 

Constructor & Destructor Documentation

mxnet::cpp::AdaGradOptimizer::AdaGradOptimizer ( unsigned  begin_num_update = 0)
explicit

Member Function Documentation

std::string mxnet::cpp::AdaGradOptimizer::GetType ( ) const
overridevirtual

get optimizer type

Returns
string of optimizer type

Implements mxnet::cpp::Optimizer.

void mxnet::cpp::AdaGradOptimizer::Update ( int  index,
NDArray  weight,
NDArray  grad 
)
overridevirtual

Update a weight with gradient.

Parameters
indexthe unique index for the weight.
weightthe weight to update.
gradgradient for the weight.

Implements mxnet::cpp::Optimizer.


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