get_dl_pruning_paramT_get_dl_pruning_paramGetDlPruningParamGetDlPruningParamget_dl_pruning_param (Operator)
Name
get_dl_pruning_paramT_get_dl_pruning_paramGetDlPruningParamGetDlPruningParamget_dl_pruning_param — Get information from a pruning data handle.
Signature
Description
get_dl_pruning_paramget_dl_pruning_paramGetDlPruningParamGetDlPruningParamget_dl_pruning_param returns the parameter and pruning values
GenParamValueGenParamValueGenParamValuegenParamValuegen_param_value of GenParamNameGenParamNameGenParamNamegenParamNamegen_param_name from the pruning data
handle DLPruningHandleDLPruningHandleDLPruningHandleDLPruningHandledlpruning_handle.
The values of following parameters GenParamNameGenParamNameGenParamNamegenParamNamegen_param_name can be retrieved:
- 'mode'"mode""mode""mode""mode":
-
Specifies the pruning method
See create_dl_pruningcreate_dl_pruningCreateDlPruningCreateDlPruningcreate_dl_pruning for further information.
Default: 'oracle'"oracle""oracle""oracle""oracle"
- 'percentage'"percentage""percentage""percentage""percentage":
-
Determines how much will be removed in
percents. E.g., in case of 'mode'"mode""mode""mode""mode"='oracle'"oracle""oracle""oracle""oracle" it specifies
the relative amount of possibly removable kernels to be removed, whereby
only removable kernels are considered.
Default: 10
Additionally, for 'mode'"mode""mode""mode""mode"='oracle'"oracle""oracle""oracle""oracle", the following pruning
values can be retrieved using GenParamNameGenParamNameGenParamNamegenParamNamegen_param_name:
- 'prunable_conv_num'"prunable_conv_num""prunable_conv_num""prunable_conv_num""prunable_conv_num":
Number of prunable convolutional layers
in the network. Whether a layer is prunable depends on the
specifications of the layer itself as well as the network architecture.
- 'prunable_conv_layer_name'"prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name":
Names of the prunable
convolutional layers in the network.
- 'prunable_kernel_num'"prunable_kernel_num""prunable_kernel_num""prunable_kernel_num""prunable_kernel_num":
Number of prunable convolution kernels
in the network. It considers that a pruned convolution has to retain at
least 1 convolution kernel.
- 'pruned_kernel_num'"pruned_kernel_num""pruned_kernel_num""pruned_kernel_num""pruned_kernel_num":
Number of pruned convolution kernels in
the network after a call of gen_dl_pruned_modelgen_dl_pruned_modelGenDlPrunedModelGenDlPrunedModelgen_dl_pruned_model.
- 'pruned_percentage'"pruned_percentage""pruned_percentage""pruned_percentage""pruned_percentage":
Actual percentage of pruned convolution
kernels in the network after a call of gen_dl_pruned_modelgen_dl_pruned_modelGenDlPrunedModelGenDlPrunedModelgen_dl_pruned_model: the
ratio 'pruned_kernel_num'"pruned_kernel_num""pruned_kernel_num""pruned_kernel_num""pruned_kernel_num" versus 'prunable_kernel_num'"prunable_kernel_num""prunable_kernel_num""prunable_kernel_num""prunable_kernel_num".
The following table gives an overview, which parameters and values can be set
using set_dl_pruning_paramset_dl_pruning_paramSetDlPruningParamSetDlPruningParamset_dl_pruning_param and which ones can be retrieved
using get_dl_pruning_paramget_dl_pruning_paramGetDlPruningParamGetDlPruningParamget_dl_pruning_param.
|
Parameters |
set |
get |
|
'mode'"mode""mode""mode""mode" |
x |
x
|
|
'percentage'"percentage""percentage""percentage""percentage" |
x |
x
|
|
'prunable_conv_num'"prunable_conv_num""prunable_conv_num""prunable_conv_num""prunable_conv_num" |
|
x
|
|
'prunable_conv_layer_name'"prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name" |
|
x
|
|
'prunable_kernel_num'"prunable_kernel_num""prunable_kernel_num""prunable_kernel_num""prunable_kernel_num" |
|
x
|
|
'pruned_kernel_num'"pruned_kernel_num""pruned_kernel_num""pruned_kernel_num""pruned_kernel_num" |
|
x
|
|
'pruned_percentage'"pruned_percentage""pruned_percentage""pruned_percentage""pruned_percentage" |
|
x
|
Execution Information
- Multithreading type: reentrant (runs in parallel with non-exclusive operators).
- Multithreading scope: global (may be called from any thread).
- Processed without parallelization.
Parameters
DLPruningHandleDLPruningHandleDLPruningHandleDLPruningHandledlpruning_handle (input_control) dl_pruning → HDlPrune, HTupleHHandleHTupleHtuple (handle) (IntPtr) (HHandle) (handle)
Pruning data handle.
GenParamNameGenParamNameGenParamNamegenParamNamegen_param_name (input_control) attribute.name → HTuplestrHTupleHtuple (string) (string) (HString) (char*)
Name of the generic parameter.
Default:
'percentage'
"percentage"
"percentage"
"percentage"
"percentage"
List of values:
'mode'"mode""mode""mode""mode", 'percentage'"percentage""percentage""percentage""percentage", 'prunable_conv_layer_name'"prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name""prunable_conv_layer_name", 'prunable_conv_num'"prunable_conv_num""prunable_conv_num""prunable_conv_num""prunable_conv_num", 'prunable_kernel_num'"prunable_kernel_num""prunable_kernel_num""prunable_kernel_num""prunable_kernel_num", 'pruned_kernel_num'"pruned_kernel_num""pruned_kernel_num""pruned_kernel_num""pruned_kernel_num", 'pruned_percentage'"pruned_percentage""pruned_percentage""pruned_percentage""pruned_percentage"
GenParamValueGenParamValueGenParamValuegenParamValuegen_param_value (output_control) attribute.name(-array) → HTupleSequence[Union[str, float, int]]HTupleHtuple (integer / string / real) (int / long / string / double) (Hlong / HString / double) (Hlong / char* / double)
Value of the generic parameter.
Possible Predecessors
create_dl_pruningcreate_dl_pruningCreateDlPruningCreateDlPruningcreate_dl_pruning,
set_dl_pruning_paramset_dl_pruning_paramSetDlPruningParamSetDlPruningParamset_dl_pruning_param
See also
set_dl_pruning_paramset_dl_pruning_paramSetDlPruningParamSetDlPruningParamset_dl_pruning_param
Module
Deep Learning Training