Class that computes the results of the backward dropout layer in the batch processing mode.
More...
◆ DropoutBackwardBatch() [1/2]
Constructs the backward dropout layer by copying input objects of backward dropout layer
- Parameters
-
context | Context to manage the backward dropout layer |
other | A backward dropout layer to be used as the source to initialize the input objects of the backward dropout layer |
◆ DropoutBackwardBatch() [2/2]
Constructs the backward dropout layer
- Parameters
-
context | Context to manage the backward dropout layer |
cls | Data type to use in intermediate computations for the layer, Double.class or Float.class |
method | The layer computation method, DropoutMethod |
◆ clone()
Returns the newly allocated backward dropout layer with a copy of input objects of this backward dropout layer
- Parameters
-
context | Context to manage the layer |
- Returns
- The newly allocated backward dropout layer
◆ compute()
Computes the result of the backward dropout layer
- Returns
- Backward dropout layer result
◆ getLayerInput()
Returns the structure that contains input object of the backward layer
- Returns
- Structure that contains input object of the backward layer
◆ getLayerParameter()
Returns the structure that contains parameters of the backward layer
- Returns
- Structure that contains parameters of the backward layer
◆ getLayerResult()
Returns the structure that contains result of the backward layer
- Returns
- Structure that contains result of the backward layer
◆ setResult()
Registers user-allocated memory to store the result of the backward dropout layer
- Parameters
-
result | Structure to store the result of the backward dropout layer |
◆ input
◆ method
Computation method for the layer
◆ parameter
DropoutParameters of the layer
The documentation for this class was generated from the following file:
- DropoutBackwardBatch.java