TypeError: 'NoneType' object is not callable

[liaoyanchun]

Hi!
I meet the following error while training. Can anyone help me?

2022-02-14 14:33:01,692 INFO    [offset_helper.py, 54] engery/max-distance: 5 engery/min-distance: 0
2022-02-14 14:33:01,692 INFO    [offset_helper.py, 61] direction/num_classes: 8 scale: 1
2022-02-14 14:33:01,692 INFO    [offset_helper.py, 66] c4 align axis: False
2022-02-14 14:33:01,711 INFO    [module_runner.py, 44] BN Type is inplace_abn.
2022-02-14 14:33:01,711 INFO    [__init__.py, 17] Using evaluator: StandardEvaluator
Traceback (most recent call last):
  File "main.py", line 215, in <module>
    model = Trainer(configer)
  File "D:\openseg\segmentor\trainer.py", line 70, in __init__
    self._init_model()
  File "D:\openseg\segmentor\trainer.py", line 73, in _init_model
    self.seg_net = self.model_manager.semantic_segmentor()
  File "D:\openseg\lib\models\model_manager.py", line 86, in semantic_segmentor
    model = SEG_MODEL_DICT[model_name](self.configer)
  File "D:\openseg\lib\models\nets\hrnet.py", line 105, in __init__
    self.backbone = BackboneSelector(configer).get_backbone()
  File "D:\openseg\lib\models\backbones\backbone_selector.py", line 34, in get_backbone
    model = HRNetBackbone(self.configer)(**params)
  File "D:\openseg\lib\models\backbones\hrnet\hrnet_backbone.py", line 769, in __call__
    bn_momentum=0.1)
  File "D:\openseg\lib\models\backbones\hrnet\hrnet_backbone.py", line 313, in __init__
    self.bn1 = ModuleHelper.BatchNorm2d(bn_type=bn_type)(64, momentum=bn_momentum)
TypeError: 'NoneType' object is not callable

I followed the guidelines GETTING_STARTED.md and run bash scripts/cityscapes/hrnet/run_h_48_d_4_ocr.sh train 1 . I have not changed any code in configs/cityscapes/H_48_D_4.json and scripts/cityscapes/hrnet/run_h_48_d_4_ocr.sh

[liaoyanchun]

I tried to use the trained model to evaluate HRNet-W48 + OCR on Cityscapes(guided by MODEL_ZOO.md), and this error occured again.

2022-02-14 20:05:14,497 INFO    [offset_helper.py, 54] engery/max-distance: 5 engery/min-distance: 0
2022-02-14 20:05:14,497 INFO    [offset_helper.py, 61] direction/num_classes: 8 scale: 1
2022-02-14 20:05:14,497 INFO    [offset_helper.py, 66] c4 align axis: False
2022-02-14 20:05:14,557 INFO    [module_runner.py, 44] BN Type is inplace_abn.
Traceback (most recent call last):
  File "main.py", line 218, in <module>
    model = Tester(configer)
  File "D:\openseg\segmentor\tester.py", line 69, in __init__
    self._init_model()
  File "D:\openseg\segmentor\tester.py", line 72, in _init_model
    self.seg_net = self.model_manager.semantic_segmentor()
  File "D:\openseg\lib\models\model_manager.py", line 86, in semantic_segmentor
    model = SEG_MODEL_DICT[model_name](self.configer)
  File "D:\openseg\lib\models\nets\hrnet.py", line 105, in __init__
    self.backbone = BackboneSelector(configer).get_backbone()
  File "D:\openseg\lib\models\backbones\backbone_selector.py", line 34, in get_backbone
    model = HRNetBackbone(self.configer)(**params)
  File "D:\openseg\lib\models\backbones\hrnet\hrnet_backbone.py", line 769, in __call__
    bn_momentum=0.1)
  File "D:\openseg\lib\models\backbones\hrnet\hrnet_backbone.py", line 313, in __init__
    self.bn1 = ModuleHelper.BatchNorm2d(bn_type=bn_type)(64, momentum=bn_momentum)
TypeError: 'NoneType' object is not callable

Help me....

[liaoyanchun]

I solved the problem by changing the version of PyTorch(from 1.8.0 to 1.0.0), inspired by this post.