Interface | Description |
---|---|
Classifier |
A classifier (this is for supervised learning)
|
Layer |
Interface for a layer of a neural network.
|
Model |
A Model is meant for predicting something from data.
|
NeuralNetwork | |
ParamInitializer |
Param initializer for a layer
|
Trainable |
Trainable: an interface common to Layers and GraphVertices that have trainable parameters
|
TrainingConfig |
Simple interface for the training configuration (updater, L1/L2 values, etc) for trainable layers/vertices.
|
Updater |
Update the model
|
Enum | Description |
---|---|
FwdPassType |
Type of forward pass to do.
|
Layer.TrainingMode | |
Layer.Type | |
MaskState |
MaskState: specifies whether a mask should be applied or not.
|
OptimizationAlgorithm |
Optimization algorithm to use
|
Copyright © 2018. All rights reserved.