Hello, I am also working on this issue now, what method did you use. I replaced the encoder with timm-regnet, but the network will report the following error, please also point me out
Traceback (most recent call last):
File "D:\HybridNets-main\train.py", line 320, in
train(opt)
File "D:\mypyfile\HybridNets-main-revise\train.py", line 152, in train
model = HybridNetsBackbone(num_classes=len(params.obj_list), compound_coef=opt.compound_coef,
File "D:\HybridNets-main\backbone.py", line 95, in init
self.encoder = get_encoder( # 'efficientnet-b' + str(self.backbone_compound_coef[compound_coef])
File "D:\HybridNets-main\encoders__init.py", line 67, in get_encoder
encoder = Encoder(**params)
File "D:\HybridNets-main\encoders\timm_regnet.py", line 8, in init
super().init(**kwargs)
File "D:\miniconda\envs\yo\lib\site-packages\timm\models\regnet.py", line 389, in init__
stem_width = cfg.stem_width
AttributeError: 'dict' object has no attribute 'stem_width'
Hello, I am also working on this issue now, what method did you use. I replaced the encoder with timm-regnet, but the network will report the following error, please also point me out
Traceback (most recent call last): File "D:\HybridNets-main\train.py", line 320, in
train(opt)
File "D:\mypyfile\HybridNets-main-revise\train.py", line 152, in train
model = HybridNetsBackbone(num_classes=len(params.obj_list), compound_coef=opt.compound_coef,
File "D:\HybridNets-main\backbone.py", line 95, in init
self.encoder = get_encoder( # 'efficientnet-b' + str(self.backbone_compound_coef[compound_coef])
File "D:\HybridNets-main\encoders__init.py", line 67, in get_encoder
encoder = Encoder(**params)
File "D:\HybridNets-main\encoders\timm_regnet.py", line 8, in init
super().init(**kwargs)
File "D:\miniconda\envs\yo\lib\site-packages\timm\models\regnet.py", line 389, in init__
stem_width = cfg.stem_width
AttributeError: 'dict' object has no attribute 'stem_width'