Add maximum pruning rate statistics (openvinotoolkit#931)

Add maximum pruning rate statistics/count of prunable layers which was actually pruned

nncf/common/pruning/statistics.py

@@ -98,10 +98,10 @@ def __init__(self,
         Initializes statistics of the filter pruning algorithm.
 
         :param model_statistics: Statistics of the pruned model.
-        :param full_flops: Full FLOPS.
-        :param current_flops: Current FLOPS.
-        :param full_params_num: Full number of weights.
-        :param current_params_num: Current number of weights.
+        :param full_flops: The total amount of FLOPS in the model.
+        :param current_flops: Current amount of FLOPS in the model.
+        :param full_params_num: The total amount of weights in the model.
+        :param current_params_num: Current amount of weights in the model.
         :param target_pruning_level: A target level of the pruning
             for the algorithm for the current epoch.
         """
@@ -133,3 +133,57 @@ def to_str(self) -> str:
             f'Statistics of the filter pruning algorithm:\n{algorithm_string}'
         )
         return pretty_string
+
+
+class PrunedModelTheoreticalBorderline(Statistics):
+    """
+    Contains theoretical borderline statistics of the filter pruning algorithm.
+    """
+
+    def __init__(self,
+                 num_pruned_layers: int,
+                 num_prunable_layers: int,
+                 max_prunable_flops: float,
+                 max_prunable_params: float,
+                 total_flops: int,
+                 total_params: int):
+        """
+        Initializes statistics of the filter pruning theoretical borderline.
+
+        :param num_pruned_layers: Number of layers which was actually
+            pruned.
+        :param num_prunable_layers: Number of layers which have
+            prunable type.
+        :param max_prunable_flops: Number of flops for pruned
+            model with pruning rate = 1.
+        :param max_prunable_params: Number of weights for pruned
+            model with pruning rate = 1.
+        :param total_flops: The total amount of FLOPS in the model.
+        :param total_params: The total amount of weights in the model.
+        """
+        self._giga = 1e9
+        self._mega = 1e6
+        self.pruned_layers_num = num_pruned_layers
+        self.prunable_layers_num = num_prunable_layers
+        self.minimum_possible_flops = max_prunable_flops
+        self.minimum_possible_params = max_prunable_params
+        self.total_flops = total_flops
+        self.total_params = total_params
+
+    def to_str(self) -> str:
+        algorithm_string = create_table(
+            header=['Statistic\'s name', 'Value'],
+            rows=[
+                ['Pruned layers count / prunable layers count', f'{self.pruned_layers_num} /'
+                                                                f' {self.prunable_layers_num}'],
+                ['GFLOPS minimum possible after pruning / total', f'{self.minimum_possible_flops / self._giga:.3f} /'
+                                                                  f' {self.total_flops / self._giga:.3f}'],
+                ['MParams minimum possible after pruning / total', f'{self.minimum_possible_params / self._mega:.3f} /'
+                                                                   f' {self.total_params / self._mega:.3f}'],
+            ]
+        )
+
+        pretty_string = (
+            f'Theoretical borderline of the filter pruning algorithm\nfor current model:\n{algorithm_string}'
+        )
+        return pretty_string

nncf/tensorflow/pruning/filter_pruning/algorithm.py

@@ -34,6 +34,8 @@
 from nncf.common.pruning.utils import get_conv_in_out_channels
 from nncf.common.pruning.utils import get_rounded_pruned_element_number
 from nncf.common.statistics import NNCFStatistics
+from nncf.common.pruning.statistics import PrunedModelTheoreticalBorderline
+from nncf.common.utils.debug import is_debug
 from nncf.common.utils.logger import logger as nncf_logger
 from nncf.common.schedulers import StubCompressionScheduler
 from nncf.common.accuracy_aware_training.training_loop import ADAPTIVE_COMPRESSION_CONTROLLERS
@@ -126,6 +128,10 @@ def __init__(self,
         self.current_flops = self.full_flops
         self.full_params_num = sum(self._nodes_params_num.values())
         self.current_params_num = self.full_params_num
+        self._pruned_layers_num = len(self._pruned_layer_groups_info.get_all_nodes())
+        self._prunable_layers_num = len(self._original_graph.get_nodes_by_types(self._prunable_types))
+        self._max_prunable_flops, self._max_prunable_params = \
+            self._calculate_flops_and_weights_in_uniformly_pruned_model(1.)
 
         self._weights_normalizer = tensor_l2_normalizer  # for all weights in common case
         self._filter_importance = FILTER_IMPORTANCE_FUNCTIONS.get(params.get('filter_importance', 'L2'))
@@ -162,6 +168,13 @@ def disable_scheduler(self):
         self._scheduler.current_pruning_level = 0.0
 
     def statistics(self, quickly_collected_only: bool = False) -> NNCFStatistics:
+        if not quickly_collected_only and is_debug():
+            stats = PrunedModelTheoreticalBorderline(
+                self._pruned_layers_num, self._prunable_layers_num, self._max_prunable_flops,
+                self._max_prunable_params, self.full_flops, self.full_params_num)
+
+            nncf_logger.debug(stats.to_str())
+
         model_statistics = self._calculate_pruned_model_stats()
         self._update_benchmark_statistics()
         target_pruning_level = self.scheduler.current_pruning_level

nncf/torch/pruning/filter_pruning/algo.py

@@ -31,6 +31,7 @@
 from nncf.common.pruning.mask_propagation import MaskPropagationAlgorithm
 from nncf.common.pruning.schedulers import PRUNING_SCHEDULERS
 from nncf.common.pruning.statistics import FilterPruningStatistics
+from nncf.common.pruning.statistics import PrunedModelTheoreticalBorderline
 from nncf.common.pruning.statistics import PrunedLayerSummary
 from nncf.common.pruning.statistics import PrunedModelStatistics
 from nncf.common.pruning.utils import calculate_in_out_channels_in_uniformly_pruned_model
@@ -41,6 +42,7 @@
 from nncf.common.pruning.utils import get_rounded_pruned_element_number
 from nncf.common.schedulers import StubCompressionScheduler
 from nncf.common.statistics import NNCFStatistics
+from nncf.common.utils.debug import is_debug
 from nncf.common.utils.logger import logger as nncf_logger
 from nncf.config.extractors import extract_bn_adaptation_init_params
 from nncf.torch.algo_selector import PT_COMPRESSION_ALGORITHMS
@@ -118,6 +120,7 @@ def __init__(self, target_model: NNCFNetwork,
                  prunable_types: List[str],
                  pruned_module_groups: Clusterization[PrunedModuleInfo],
                  config: NNCFConfig):
+        #pylint:disable=too-many-statements
         super().__init__(target_model, prunable_types, pruned_module_groups, config)
         params = self.pruning_config.get('params', {})
         self.frozen = False
@@ -137,6 +140,10 @@ def __init__(self, target_model: NNCFNetwork,
         self.current_flops = self.full_flops
         self.full_params_num = sum(self.nodes_params_num.values())
         self.current_params_num = self.full_params_num
+        self._pruned_layers_num = len(self.pruned_module_groups_info.get_all_nodes())
+        self._prunable_layers_num = len(self._model.get_graph().get_nodes_by_types(self._prunable_types))
+        self._max_prunable_flops, self._max_prunable_params =\
+            self._calculate_flops_and_weights_in_uniformly_pruned_model(1.)
 
         self.weights_normalizer = tensor_l2_normalizer  # for all weights in common case
         self.filter_importance = FILTER_IMPORTANCE_FUNCTIONS.get(params.get('filter_importance', 'L2'))
@@ -206,6 +213,13 @@ def set_mask(minfo: PrunedModuleInfo, mask: torch.Tensor) -> None:
         minfo.operand.binary_filter_pruning_mask = mask
 
     def statistics(self, quickly_collected_only: bool = False) -> NNCFStatistics:
+        if not quickly_collected_only and is_debug():
+            stats = PrunedModelTheoreticalBorderline(
+                self._pruned_layers_num, self._prunable_layers_num, self._max_prunable_flops,
+                self._max_prunable_params, self.full_flops, self.full_params_num)
+
+            nncf_logger.debug(stats.to_str())
+
         pruned_layers_summary = {}
         for minfo in self.pruned_module_groups_info.get_all_nodes():
             layer_name = str(minfo.module_scope)