Pylint2.10 update

examples/torch/classification/main.py

@@ -21,15 +21,15 @@
 from typing import Any
 
 import torch
-import torch.backends.cudnn as cudnn
-import torch.nn as nn
+from torch.backends import cudnn
+from torch import nn
 import torch.nn.parallel
 import torch.optim
 import torch.utils.data
 import torch.utils.data.distributed
-import torchvision.datasets as datasets
-import torchvision.models as models
-import torchvision.transforms as transforms
+from torchvision import datasets
+from torchvision import models
+from torchvision import transforms
 from torch.nn.modules.loss import _Loss
 from torch.optim.lr_scheduler import ReduceLROnPlateau
 from torchvision.datasets import CIFAR10
@@ -79,7 +79,7 @@
 from nncf.torch.utils import is_main_process
 from nncf.torch.utils import safe_thread_call
 
-model_names = sorted(name for name in models.__dict__
+model_names = sorted(name for name in models.__dict__.items()
                      if name.islower() and not name.startswith("__")
                      and callable(models.__dict__[name]))
 

examples/torch/classification/models/mobilenet_v2_32x32.py

@@ -14,7 +14,7 @@
 from typing import List
 from typing import Optional
 
-import torch.nn as nn
+from torch import nn
 from torch import Tensor
 
 

examples/torch/classification/staged_quantization_worker.py

@@ -16,8 +16,8 @@
 import time
 
 import torch
-import torch.backends.cudnn as cudnn
-import torch.nn as nn
+from torch.backends import cudnn
+from torch import nn
 import torch.nn.parallel
 import torch.optim
 import torch.utils.data

examples/torch/common/model_loader.py

@@ -21,7 +21,7 @@
 import examples.torch.common.models as custom_models
 from examples.torch.classification.models.mobilenet_v2_32x32 import MobileNetV2For32x32
 from examples.torch.common.example_logger import logger
-import examples.torch.common.restricted_pickle_module as restricted_pickle_module
+from examples.torch.common import restricted_pickle_module
 from nncf.torch.checkpoint_loading import load_state
 from nncf.torch.utils import safe_thread_call
 

examples/torch/common/models/classification/inceptionv3_cifar100.py

@@ -17,7 +17,7 @@
 from collections import namedtuple
 
 import torch
-import torch.nn as nn
+from torch import nn
 import torch.nn.functional as F
 from torch.utils import model_zoo
 
@@ -94,7 +94,7 @@ def __init__(self, num_classes=1000, aux_logits=True, transform_input=False):
 
         for m in self.modules():
             if isinstance(m, (nn.Conv2d, nn.Linear)):
-                import scipy.stats as stats
+                from scipy import stats
                 stddev = m.stddev if hasattr(m, 'stddev') else 0.1
                 X = stats.truncnorm(-2, 2, scale=stddev)
                 values = torch.as_tensor(X.rvs(m.weight.numel()), dtype=m.weight.dtype)

examples/torch/common/models/classification/mobilenetv3.py

@@ -21,7 +21,7 @@
 
 """
 
-import torch.nn as nn
+from torch import nn
 import math
 
 __all__ = ['mobilenetv3_Large', 'mobilenetv3']

examples/torch/common/models/classification/resnet_cifar.py

@@ -11,7 +11,7 @@
  limitations under the License.
 """
 
-import torch.nn as nn
+from torch import nn
 import torch.nn.functional as F
 
 

examples/torch/common/models/classification/rmnet_cifar.py

@@ -1,6 +1,6 @@
 from collections import OrderedDict
 
-import torch.nn as nn
+from torch import nn
 import torch.nn.functional as F
 
 

examples/torch/common/models/segmentation/enet.py

@@ -14,7 +14,7 @@
 # ENet implementation from:
 # https://github.com/davidtvs/PyTorch-ENet
 
-import torch.nn as nn
+from torch import nn
 import torch
 
 from examples.torch.common.example_logger import logger

examples/torch/common/models/segmentation/icnet.py

@@ -22,7 +22,7 @@
 from pkg_resources import parse_version
 
 from numpy import lcm
-import torch.nn as nn
+from torch import nn
 import torch
 import torch.nn.functional as F
 

examples/torch/common/models/segmentation/unet.py

@@ -16,7 +16,7 @@
 
 from pkg_resources import parse_version
 
-import torch.nn as nn
+from torch import nn
 import torch
 import torch.nn.functional as F
 

examples/torch/common/restricted_pickle_module.py

@@ -63,7 +63,7 @@ def find_class(self, module_name, class_name):
             return getattr(builtins, class_name)
         if module_name == "collections" and class_name in Unpickler.safe_collections:
             return getattr(collections, class_name)
-        for allowed_module_name in Unpickler.allowed_classes:
+        for allowed_module_name in Unpickler.allowed_classes.items():
             if module_name == allowed_module_name and class_name in Unpickler.allowed_classes[allowed_module_name]:
                 module = importlib.import_module(module_name)
                 return getattr(module, class_name)

examples/torch/common/utils.py

@@ -25,7 +25,7 @@
 from typing import Tuple
 
 from PIL import Image
-import torch.utils.data as data
+from torch.utils import data
 
 from examples.torch.common.distributed import configure_distributed
 from examples.torch.common.execution import ExecutionMode, get_device
@@ -76,7 +76,7 @@ def get_name(config):
 def write_metrics(acc, filename):
     avg = round(acc * 100, 2)
     metrics = {"Accuracy": avg}
-    with open(filename, 'w') as outfile:
+    with open(filename, 'w', encoding='utf8') as outfile:
         json.dump(metrics, outfile)
 
 
@@ -223,8 +223,9 @@ class ForkedPdb(pdb.Pdb):
     def interaction(self, *args, **kwargs):
         _stdin = sys.stdin
         try:
-            sys.stdin = open('/dev/stdin')
-            pdb.Pdb.interaction(self, *args, **kwargs)
+            with open('/dev/stdin', encoding='utf8') as file:
+                sys.stdin = file
+                pdb.Pdb.interaction(self, *args, **kwargs)
         finally:
             sys.stdin = _stdin
 

examples/torch/object_detection/datasets/coco.py

@@ -19,7 +19,7 @@
 import cv2
 import numpy as np
 import torch
-import torch.utils.data as data
+from torch.utils import data
 
 from examples.torch.common.example_logger import logger
 
@@ -50,7 +50,7 @@ def _read_coco_annotation(annotation_file, images_folder):
     images_folder = Path(images_folder)
     anno_dict = OrderedDict()
 
-    with open(annotation_file) as data_file:
+    with open(annotation_file, encoding="utf8") as data_file:
         json_annotation = json.load(data_file)
     annotation = json_annotation["annotations"]
 

examples/torch/object_detection/datasets/voc0712.py

@@ -18,7 +18,7 @@
 import cv2
 import numpy as np
 import torch
-import torch.utils.data as data
+from torch.utils import data
 
 if sys.version_info[0] == 2:
     import defusedxml.cElementTree as ET
@@ -115,12 +115,13 @@ def __init__(self, root, image_sets=(('2007', 'trainval'), ('2012', 'trainval'))
         self.return_image_info = return_image_info
         self._annopath = os.path.join('%s', 'Annotations', '%s.xml')
         self._imgpath = os.path.join('%s', 'JPEGImages', '%s.jpg')
-        self.ids = list()
+        self.ids = []
 
         for (year, name) in self.image_set:
             rootpath = os.path.join(self.root, 'VOC' + year)
-            for line in open(os.path.join(rootpath, 'ImageSets', 'Main', name + '.txt')):
-                self.ids.append((rootpath, line.strip()))
+            with open(os.path.join(rootpath, 'ImageSets', 'Main', name + '.txt'), encoding='utf8') as lines:
+                for line in lines:
+                    self.ids.append((rootpath, line.strip()))
 
     def __getitem__(self, index):
         """

examples/torch/object_detection/eval.py

@@ -217,11 +217,11 @@ def load_detection_annotations(cachedir, dataset):
             # save
             logger.info('Saving cached annotations to {:s}'.format(cachefile))
             pathlib.Path(cachedir).mkdir(parents=True, exist_ok=True)
-            with open(cachefile, 'w') as f:
+            with open(cachefile, 'w', encoding='utf8') as f:
                 json.dump(gt, f)
     if is_dist_avail_and_initialized():
         dist.barrier()
-    with open(cachefile, 'r') as f:
+    with open(cachefile, 'r', encoding='utf8') as f:
         gt = json.load(f)
     return gt, imagenames
 

examples/torch/object_detection/layers/functions/detection.py

@@ -96,8 +96,8 @@ def forward(ctx, loc_data, conf_data, prior_data, detection_output_params):
                 conf_scores = conf_preds[i].clone()
 
                 total_detections_count = 0
-                all_indices = dict()  # indices of confident detections for each class
-                boxes = dict()
+                all_indices = {} # indices of confident detections for each class
+                boxes = {}
                 for cl in range(0, detection_output_params.num_classes):
                     if cl == detection_output_params.background_label_id:
                         continue
@@ -115,7 +115,7 @@ def forward(ctx, loc_data, conf_data, prior_data, detection_output_params):
                     all_indices[cl] = all_indices[cl][:count]
                     total_detections_count += count
 
-                score_index_pairs = list()  # list of tuples (score, label, idx)
+                score_index_pairs = []  # list of tuples (score, label, idx)
                 for label, indices in all_indices.items():
                     indices = indices.cpu().numpy()
                     for idx in indices:
@@ -124,7 +124,7 @@ def forward(ctx, loc_data, conf_data, prior_data, detection_output_params):
                 score_index_pairs.sort(key=lambda tup: tup[0], reverse=True)
                 score_index_pairs = score_index_pairs[:detection_output_params.keep_top_k]
 
-                all_indices_new = dict()
+                all_indices_new = {}
                 for _, label, idx in score_index_pairs:
                     if label not in all_indices_new:
                         all_indices_new[label] = [idx]

examples/torch/object_detection/layers/modules/l2norm.py

@@ -12,8 +12,8 @@
 """
 
 import torch
-import torch.nn as nn
-import torch.nn.init as init
+from torch import nn
+from torch.nn import init
 
 from nncf.torch.utils import add_domain
 from nncf.torch import register_module

examples/torch/object_detection/layers/modules/multibox_loss.py

@@ -12,7 +12,7 @@
 """
 
 import torch
-import torch.nn as nn
+from torch import nn
 import torch.nn.functional as F
 
 from ..box_utils import match, log_sum_exp

examples/torch/object_detection/layers/modules/ssd_head.py

@@ -106,6 +106,7 @@ def __init__(self, outputs, modules):
         super().__init__(*modules)
         self.outputs = [str(o) for o in outputs]
 
+    #pylint:disable=W0237
     def forward(self, x):
         outputs = []
         for name, module in self._modules.items():

examples/torch/object_detection/main.py

@@ -17,7 +17,7 @@
 from pathlib import Path
 
 import torch
-import torch.utils.data as data
+from torch.utils import data
 
 from examples.torch.common.argparser import parse_args
 from torch.optim.lr_scheduler import ReduceLROnPlateau

examples/torch/object_detection/models/ssd_mobilenet.py

@@ -12,7 +12,7 @@
 """
 
 import torch
-import torch.nn as nn
+from torch import nn
 
 from examples.torch.common import restricted_pickle_module
 from examples.torch.common.example_logger import logger

examples/torch/object_detection/models/ssd_vgg.py

@@ -13,7 +13,7 @@
 
 import os
 import torch
-import torch.nn as nn
+from torch import nn
 
 from examples.torch.common import restricted_pickle_module
 from examples.torch.common.example_logger import logger

nncf/common/composite_compression.py

@@ -248,7 +248,7 @@ def get_state(self) -> Dict[str, Dict[str, Any]]:
 
         :return: The composite compression controller state.
         """
-        result = dict()
+        result = {}
         for ctrl in self.child_ctrls:
             result.update(ctrl.get_state())
         return result
@@ -394,7 +394,7 @@ def get_state(self) -> Dict[str, Dict]:
 
         :return: The composite compression builder state.
         """
-        result = dict()
+        result = {}
         for builder in self.child_builders:
             result.update(builder.get_state())
         return result

nncf/common/graph/graph.py

@@ -165,7 +165,7 @@ class NNCFGraph:
 
     def __init__(self):
         self._nx_graph = nx.DiGraph()
-        self._node_id_to_key_dict = dict()
+        self._node_id_to_key_dict = {}
         self._input_nncf_nodes = {}  # type: Dict[int, NNCFNode]
         self._output_nncf_nodes = {}  # type: Dict[int, NNCFNode]
 

nncf/common/graph/operator_metatypes.py

@@ -57,7 +57,7 @@ def subtype_check(cls, metatype: Type['OperatorMetatype']) -> bool:
         if metatype == cls or metatype in subtypes:
             return True
 
-        return any([subtype.subtype_check(metatype) for subtype in subtypes])
+        return any(subtype.subtype_check(metatype) for subtype in subtypes)
 
 
 class OperatorMetatypeRegistry(Registry):

nncf/common/graph/patterns.py

@@ -33,7 +33,7 @@ class HWFusedPatterns:
     """
 
     def __init__(self):
-        self._patterns_dict = dict()
+        self._patterns_dict = {}
         self._full_pattern_graph = GraphPattern()
 
     def register(self, pattern: 'GraphPattern', name: str, match: bool = True) -> None:

nncf/common/pruning/pruning_node_selector.py

@@ -103,7 +103,7 @@ def create_pruning_groups(self, graph: NNCFGraph) -> Clusterization[NNCFNode]:
         for i, cluster in enumerate(special_ops_clusterization.get_all_clusters()):
             all_pruned_inputs = []
             pruned_inputs_idxs = set()
-            clusters_to_merge = list()
+            clusters_to_merge = []
 
             for node in cluster.elements:
                 sources = get_sources_of_node(node, graph, self._prune_operations)

nncf/common/quantization/quantizer_propagation/solver.py

@@ -140,7 +140,7 @@ def constrain_quantizer_config_list_for_insertion(self, quantization_point_id: Q
           of the quantizer configs for the quantization point defined by `quantization_point_id`.
         """
         prior_list = self.quantizer_setup.quantization_points[quantization_point_id].possible_qconfigs
-        if not all([qc in prior_list for qc in constrained_config_list]):
+        if not all(qc in prior_list for qc in constrained_config_list):
             raise RuntimeError('Constrained config list is incompatible with the result of the quantizer propagation!')
         # TODO (vshampor): only allow to constrain 'input-group'-wise?
         self.quantizer_setup.quantization_points[quantization_point_id].possible_qconfigs = constrained_config_list
@@ -1193,7 +1193,7 @@ def compatible_wo_requant(qconf: QuantizerConfig,
             raise RuntimeError("Unknown propagation strategy: {}".format(self._propagation_strategy))
 
         for qconf in qconfigs_union:
-            if all([compatible_fn(qconf, qconf_list) for qconf_list in potential_qconfigs_for_each_branch]):
+            if all(compatible_fn(qconf, qconf_list) for qconf_list in potential_qconfigs_for_each_branch):
                 merged_qconfig_list.append(qconf)
 
         nncf_logger.debug("Merged list before sorting: {}".format(";".join([str(qc) for qc in merged_qconfig_list])))

nncf/common/quantization/quantizer_setup.py

@@ -447,7 +447,7 @@ def select_qconfigs(self, qp_id_vs_selected_qconfig_dict: Dict[QuantizationPoint
         if Counter(qp_id_vs_selected_qconfig_dict.keys()) != Counter(self.quantization_points.keys()):
             raise ValueError("The set of quantization points for a selection is inconsistent with quantization"
                              "points in the quantizer setup!")
-        for qp_id in self.quantization_points:
+        for qp_id in self.quantization_points.items():
             if strict:
                 retval.quantization_points[qp_id] = self.quantization_points[qp_id].select_qconfig(
                     qp_id_vs_selected_qconfig_dict[qp_id]