Apache SINGA
A distributed deep learning platform .
 All Classes Namespaces Files Functions Variables Typedefs Macros
Public Member Functions | Static Public Member Functions | Protected Attributes | List of all members
singa::LRGenerator Class Reference

Base learning rate generator. More...

#include <updater.h>

Inheritance diagram for singa::LRGenerator:
singa::ExpLRGen singa::FixedStepLRGen singa::InvLRGen singa::InvTLRGen singa::LinearLRGen singa::StepLRGen

Public Member Functions

virtual void Init (const LRGenProto &proto)
 
virtual float Get (int step)
 

Static Public Member Functions

static LRGeneratorCreate (const LRGenProto &proto)
 

Protected Attributes

LRGenProto proto_
 

Detailed Description

Base learning rate generator.

Generate learning rate for a give training step/iteration. There are many different ways to change the learning rate through time/step. Users can inherint this class to implement their own change method.

Member Function Documentation

virtual float singa::LRGenerator::Get ( int  step)
inlinevirtual
Parameters
steptraining step/iteration.
Returns
base learning rate regardless of step

Reimplemented in singa::InvTLRGen, singa::InvLRGen, singa::ExpLRGen, singa::LinearLRGen, singa::StepLRGen, and singa::FixedStepLRGen.


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