ExponentialLRDecay Class
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
This class implements Exponential Learning rate decay. Implemented from the tensorflow documentation. Source: https://www.tensorflow.org/api_docs/python/tf/compat/v1/train/exponential_decay Default values and implementation of learning rate is from Tensorflow Slim model tests. Source : https://github.com/tensorflow/models/blob/master/research/slim/train_image_classifier.py
public sealed class ExponentialLRDecay : Microsoft.ML.Trainers.LearningRateScheduler
type ExponentialLRDecay = class
inherit LearningRateScheduler
Public NotInheritable Class ExponentialLRDecay
Inherits LearningRateScheduler
- Inheritance
Constructors
ExponentialLRDecay(Single, Single, Single, Boolean) |
This contructor initializes intial learning rate, number epochs per decay, decay rate and the staircase option. The defaults are taken from Tensorflow Slim. |
Fields
DecayRate |
Learning rate decay factor. |
DecaySteps |
Number of decay steps |
GlobalStep |
The number of batches seen by the graph so far. |
LearningRate |
Initial learning rate. |
NumEpochsPerDecay |
Number of epochs after which learning rate decays. |
Staircase |
If Staircase is True the learning rate decays at discrete intervals and the decayed learning rate follows a staircase function. |