add_dl_pruning_batch🔗
Short description🔗
add_dl_pruning_batch — Calculate scores to prune a deep learning model.
Signature🔗
add_dl_pruning_batch( dl_model DLModelHandleToPrune, dl_pruning DLPruningHandle, dict DLSampleBatch )
Description🔗
add_dl_pruning_batch calculates pruning scores for the deep
learning model DLModelHandleToPrune.
More precisely, the scores are calculated on the images given in
DLSampleBatch and internally accumulated by each call of
add_dl_pruning_batch in the pruning data handle
DLPruningHandle.
The parameter DLModelHandleToPrune specifies the deep learning
model to use.
Note that add_dl_pruning_batch supports only deep learning
models of type 'classification'.
The parameter DLPruningHandle specifies the pruning data handle,
which is used to pass information as e.g., the accumulated scores or the
pruning mode.
See create_dl_pruning for further information about implemented
pruning modes.
The parameter DLSampleBatch specifies the batch with input images
based on which the scores are calculated.
Note that the number of images in the tuple DLSampleBatch needs to
be equal to the value set for the model parameter 'batch_size'.
For an explanation of the concept of deep learning see the introduction of chapter Deep Learning.
Execution information🔗
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🔗
DLModelHandleToPrune (input_control) dl_model → (handle)
Handle of a deep learning model to prune.
DLPruningHandle (input_control) dl_pruning → (handle)
Pruning data handle.
DLSampleBatch (input_control) dict-array → (handle)
Tuple of dictionaries with input images.
Combinations with other operators🔗
Combinations
Possible predecessors
read_dl_model, create_dl_pruning, set_dl_pruning_param
Possible successors
Module🔗
Deep Learning Professional