| Package | Description |
|---|---|
| org.deeplearning4j.arbiter | |
| org.deeplearning4j.arbiter.layers | |
| org.deeplearning4j.arbiter.layers.fixed |
| Modifier and Type | Field and Description |
|---|---|
protected LayerSpace<?> |
BaseNetworkSpace.LayerConf.layerSpace |
| Modifier and Type | Method and Description |
|---|---|
LayerSpace<?> |
MultiLayerSpace.getLayerSpace(int layerNumber) |
| Modifier and Type | Class and Description |
|---|---|
class |
AbstractLSTMLayerSpace<T extends org.deeplearning4j.nn.conf.layers.AbstractLSTM>
Layer space for LSTM layers
|
class |
ActivationLayerSpace
Layer space for
ActivationLayer |
class |
AutoEncoderLayerSpace
Layer space for autoencoder layers
|
class |
BaseConvolutionLayerSpace<T extends org.deeplearning4j.nn.conf.layers.FeedForwardLayer>
Layer space for convolutional layers
|
class |
BaseLayerSpace<L extends org.deeplearning4j.nn.conf.layers.BaseLayer>
BaseLayerSpace contains the common Layer hyperparameters; should match
BaseLayer in terms of features |
class |
BaseOutputLayerSpace<L extends org.deeplearning4j.nn.conf.layers.BaseOutputLayer> |
class |
BasePretrainNetworkLayerSpace<L extends org.deeplearning4j.nn.conf.layers.BasePretrainNetwork> |
class |
BatchNormalizationSpace
LayerSpace for batch normalization layers
|
class |
Bidirectional
Bidirectional layer wrapper.
|
class |
CenterLossOutputLayerSpace |
class |
ConvolutionLayerSpace
Layer space for convolutional layers
|
class |
Deconvolution2DLayerSpace |
class |
DenseLayerSpace
layer hyperparameter configuration space for dense layers (i.e., multi-layer perceptron layers)
|
class |
DropoutLayerSpace |
class |
EmbeddingLayerSpace
Layer hyperparameter configuration space for
EmbeddingLayer |
class |
FeedForwardLayerSpace<L extends org.deeplearning4j.nn.conf.layers.FeedForwardLayer> |
class |
GlobalPoolingLayerSpace
Layer space for a
GlobalPoolingLayer |
class |
GravesBidirectionalLSTMLayerSpace
Layer space for Bidirectional LSTM layers
|
class |
GravesLSTMLayerSpace
Layer space for LSTM layers
|
class |
LocalResponseNormalizationLayerSpace |
class |
LossLayerSpace |
class |
LSTMLayerSpace
Layer space for LSTM layers
|
class |
OCNNLayerSpace |
class |
OutputLayerSpace
Layer hyperparameter configuration space for output layers
|
class |
RnnOutputLayerSpace
Layer hyperparametor configuration space for RnnOutputLayer
|
class |
SeparableConvolution2DLayerSpace |
class |
SubsamplingLayerSpace
Layer hyperparameter configuration space for subsampling layers
|
class |
VariationalAutoencoderLayerSpace
Layer space for
VariationalAutoencoder |
| Modifier and Type | Field and Description |
|---|---|
protected LayerSpace<?> |
Bidirectional.layerSpace |
| Modifier and Type | Method and Description |
|---|---|
abstract <E extends LayerSpace> |
LayerSpace.Builder.build() |
<E extends LayerSpace> |
VariationalAutoencoderLayerSpace.Builder.build() |
| Constructor and Description |
|---|
Bidirectional(LayerSpace<?> layerSpace) |
| Modifier and Type | Class and Description |
|---|---|
class |
FixedLayerSpace<T extends org.deeplearning4j.nn.conf.layers.Layer>
A layer space that wraps a DL4J layer, without any optimizable hyperparameters
|
Copyright © 2020. All rights reserved.