Package | Description |
---|---|
org.deeplearning4j.nn.adapters | |
org.deeplearning4j.nn.api | |
org.deeplearning4j.nn.graph | |
org.deeplearning4j.nn.multilayer |
Modifier and Type | Class and Description |
---|---|
class |
ArgmaxAdapter
This OutputAdapter implementation is suited for silent conversion of 2D SoftMax output
|
class |
Regression2dAdapter
This OutputAdapter implementation takes single 2D nn output in, and returns JVM double[][] array
|
class |
YoloModelAdapter
This ModelAdapter implementation is suited for use of Yolo2 model with ParallelInference
|
Modifier and Type | Interface and Description |
---|---|
interface |
ModelAdapter<T>
This interface describes abstraction that uses provided model to convert INDArrays to some specific output
|
Modifier and Type | Method and Description |
---|---|
<T> T |
ComputationGraph.output(INDArray[] inputs,
INDArray[] inputMasks,
INDArray[] labelMasks,
OutputAdapter<T> outputAdapter)
This method uses provided OutputAdapter to return custom object built from INDArray
PLEASE NOTE: This method uses dedicated Workspace for output generation to avoid redundant allocations
|
Modifier and Type | Method and Description |
---|---|
<T> T |
MultiLayerNetwork.output(INDArray inputs,
INDArray inputMasks,
INDArray labelMasks,
OutputAdapter<T> outputAdapter)
This method uses provided OutputAdapter to return custom object built from INDArray
PLEASE NOTE: This method uses dedicated Workspace for output generation to avoid redundant allocations
|
Copyright © 2019. All rights reserved.