NVIDIA NvNeural SDK
2022.2
GPU inference framework for NVIDIA Nsight Deep Learning Designer
|
"Host-native" container for LayerDesc structures from other modules. More...
#include <LayerDescHostContainer.h>
Public Member Functions | |
const LayerDesc & | get () const |
Returns a reference to the converted/upgraded descriptor. | |
bool | isValid () const |
Returns true if the imported descriptor was valid. | |
LayerDescHostContainer (const LayerDesc *pInputLayerDesc) | |
Creates a LayerDescHostContainer to represent an external LayerDesc. More... | |
void | linkVectors () |
Once all sub-descriptors have been converted, call this to reconnect the parameter/input lists. | |
"Host-native" container for LayerDesc structures from other modules.
When loading a plugin that may have been built against an earlier version of the NvNeural SDK, this class creates an upgraded LayerDesc that replaces missing fields with sensible defaults.
|
explicit |
Creates a LayerDescHostContainer to represent an external LayerDesc.
pInputLayerDesc | Descriptor to convert |