search

gangweiX / IGEV-plusplus

IGEV++: Iterative Multi-range Geometry Encoding Volumes for Stereo Matching
MIT License
103 stars 5 forks source link

AttributeError: 'EfficientNetFeatures' object has no attribute 'act1' #1

Closed qqc22111 closed 2 months ago

qqc22111 commented 2 months ago

The problem description is below, how should I resolve this problem? Unexpected keys (bn2.bias, bn2.num_batches_tracked, bn2.running_mean, bn2.running_var, bn2.weight, classifier.bias, classifier.weight, conv_head.weight) found while loading pretrained weights. This may be expected if model is being adapted. Traceback (most recent call last): File "D:\Project\BinocularStereoVision\IGEV-plusplus\demo_imgs.py", line 90, in demo(args) File "D:\Project\BinocularStereoVision\IGEV-plusplus\demo_imgs.py", line 30, in demo model = torch.nn.DataParallel(IGEVStereo(args), device_ids=[0]) File "D:\Project\BinocularStereoVision\IGEV-plusplus\core\igev_stereo.py", line 107, in init self.feature = Feature() File "D:\Project\BinocularStereoVision\IGEV-plusplus\core\extractor.py", line 333, in init self.act1 = model.act1 File "D:\Anaconda\Anaconda3\envs\MedSAM\lib\site-packages\torch\nn\modules\module.py", line 1614, in getattr raise AttributeError("'{}' object has no attribute '{}'".format( AttributeError: 'EfficientNetFeatures' object has no attribute 'act1'

gangweiX commented 2 months ago

pip install timm==0.5.4

gangweiX commented 2 months ago

please install the corresponding timm version.