C++ API Reference for Intel® Data Analytics Acceleration Library 2019 Update 5
Parameters for the forward and backward pooling layers. More...
| Parameter | ( | size_t | index, |
| size_t | kernelSize = 2, |
||
| size_t | stride = 2, |
||
| size_t | padding = 0 |
||
| ) |
Constructs the parameters of 1D pooling layer
| [in] | index | Index of the dimension on which pooling is performed |
| [in] | kernelSize | Size of 1D subtensor for which the element is computed |
| [in] | stride | Interval over the dimension on which the pooling is performed |
| [in] | padding | Number of data elements to implicitly add to the the dimension of the 1D subtensor on which the pooling is performed |
| KernelSize kernelSize |
Data structure representing the size of the 1D subtensor from which the element is computed
| Padding padding |
Data structure representing the number of data elements to implicitly add to each size of the 1D subtensor on which pooling is performed
| Stride stride |
Data structure representing the intervals on which the subtensors for pooling are computed
For more complete information about compiler optimizations, see our Optimization Notice.