请问这个报错：AttributeError: 'EfficientFormerClsHead' object has no attribute 'post_process'，没有post_process()方法？

[couleur123123]

No description provided.

[Fafa-DL]

报错信息帖完整~

[couleur123123]

Loading D:\PyProj\pretrain\efficientformer-l3_3rdparty_in1k_20220803-dde1c8c5.pth
The model and loaded state dict do not match exactly

size mismatch for head.head.weight: copying a param with shape torch.Size([1000, 512]) from checkpoint, the shape in current model is torch.Size([4, 512]).
size mismatch for head.head.bias: copying a param with shape torch.Size([1000]) from checkpoint, the shape in current model is torch.Size([4]).
size mismatch for head.dist_head.weight: copying a param with shape torch.Size([1000, 512]) from checkpoint, the shape in current model is torch.Size([4, 512]).
size mismatch for head.dist_head.bias: copying a param with shape torch.Size([1000]) from checkpoint, the shape in current model is torch.Size([4]).
unexpected key in source state_dict: backbone.network.1.0.proj.weight, backbone.network.1.0.proj.bias, backbone.network.1.0.norm.weight, backbone.network.1.0.norm.bias, backbone.network.1.0.norm.running_mean, backbone.network.1.0.norm.running_var, backbone.network.1.0.norm.num_batches_tracked, backbone.network.2.0.proj.weight, backbone.network.2.0.proj.bias, backbone.network.2.0.norm.weight, backbone.network.2.0.norm.bias, backbone.network.2.0.norm.running_mean, backbone.network.2.0.norm.running_var, backbone.network.2.0.norm.num_batches_tracked, backbone.network.3.0.proj.weight, backbone.network.3.0.proj.bias, backbone.network.3.0.norm.weight, backbone.network.3.0.norm.bias, backbone.network.3.0.norm.running_mean, backbone.network.3.0.norm.running_var, backbone.network.3.0.norm.num_batches_tracked

missing keys in source state_dict: backbone.network.1.0.conv.weight, backbone.network.1.0.conv.bias, backbone.network.1.0.bn.weight, backbone.network.1.0.bn.bias, backbone.network.1.0.bn.running_mean, backbone.network.1.0.bn.running_var, backbone.network.2.0.conv.weight, backbone.network.2.0.conv.bias, backbone.network.2.0.bn.weight, backbone.network.2.0.bn.bias, backbone.network.2.0.bn.running_mean, backbone.network.2.0.bn.running_var, backbone.network.3.0.conv.weight, backbone.network.3.0.conv.bias, backbone.network.3.0.bn.weight, backbone.network.3.0.bn.bias, backbone.network.3.0.bn.running_mean, backbone.network.3.0.bn.running_var

Train: Epoch 1/100: 0%| | 0/651 [00:14<?, ?it/s<class 'dict'>]
Traceback (most recent call last):
File "D:\PyProj\Learn\Awesome-Backbones-main\utils\train_utils.py", line 221, in train
preds, losses = model(images, targets=targets, return_loss=True, train_statu=True)
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\parallel\data_parallel.py", line 183, in forward
return self.module(*inputs[0], **module_kwargs[0])
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "D:\PyProj\Learn\Awesome-Backbones-main\models\build.py", line 132, in forward
return self.forward_test(x), self.forward_train(x, **kwargs)
File "D:\PyProj\Learn\Awesome-Backbones-main\models\build.py", line 143, in forward_test
out = self.head.simple_test(x, **kwargs)
File "D:\PyProj\Learn\Awesome-Backbones-main\configs\heads\efficientformer_head.py", line 81, in simple_test
return self.post_process(pred)
File "E:\SoftwareInstall\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\module.py", line 1688, in getattr
raise AttributeError(f"'{type(self).name}' object has no attribute '{name}'")
AttributeError: 'EfficientFormerClsHead' object has no attribute 'post_process'

[couleur123123]

请问为什么一开始模型就匹配不上？

[Fafa-DL]

请问为什么一开始模型就匹配不上？

因为你的分类是4类而预训练模型是1000类，分类头肯定是不匹配的，可以理解为此处加载预训练模型更多的用他的backbone而重训分类头

[couleur123123]

刚开始接触不太明白，请问一下我尝试其他模型的时候用LinearClsHead这个分类头，我numclass也是直接改成4的，为什么这个分类头是可以直接匹配的？

[Fafa-DL]

刚开始接触不太明白，请问一下我尝试其他模型的时候用LinearClsHead这个分类头，我numclass也是直接改成4的，为什么这个分类头是可以直接匹配的？

你改了结构当然权重匹配不上啊