train error

hello,i meet a error when i run python train.py --dataset Synapse --vit_name R50-ViT-B_16 Error is as follows:

Traceback (most recent call last):
  File "train.py", line 91, in <module>
    net.load_from(weights=np.load(config_vit.pretrained_path))
  File "D:\Python\papers\TransUNet\networks\vit_seg_modeling.py", line 429, in load_from
    unit.load_from(weights, n_block=uname)
  File "D:\Python\papers\TransUNet\networks\vit_seg_modeling.py", line 192, in load_from
    query_weight = np2th(weights[pjoin(ROOT, ATTENTION_Q, "kernel")]).view(self.hidden_size, self.hidden_size).t()
  File "C:\ProgramData\Anaconda3\envs\transnet\lib\site-packages\numpy\lib\npyio.py", line 260, in __getitem__
    raise KeyError("%s is not a file in the archive" % key)
KeyError: 'Transformer/encoderblock_0\\MultiHeadDotProductAttention_1/query\\kernel is not a file in the archive'

env is corrent.and i use win10 and download R50+ViT-B_16.npz model in the right place.can you help me ?

[liyiersan]

It seems that you are not successfully access the key in the pretrained weight. Could you please just replace line 192 of vit_seg_modeling.py: query_weight = np2th(weights[pjoin(ROOT, ATTENTION_Q, “kernel”)]).view(self.hidden_size, self.hidden_size).t() with query_weight = np2th(weights[“Transformer/encoderblock_0/MultiHeadDotProductAttention_1/query/kernel”]).view(self.hidden_size, self.hidden_size).t()

If do so and there is no bug in line 192, I guess it is the problem that running on windows.

Let me know if I was wrong. Thanks.

@775397252 @YUUUCC @cwpl I found a better way to solve this. That is :

1. Create a new file path_util.py in directory networks.

   from os.path import join
   from os.path import normpath
   import platform
   
   def pjoin(path, *paths):
       p = join(path, *paths)
       if platform.system() == "Windows":
           return normpath(p).replace('\\','/')
       else:
           return p
   

2. Import pjoin in file vit_seg_modeling_resnet_skip.py and vit_seg_modeling.py

from .path_util import pjoin

1. And this may due to ‘\’ is not recognized by numpy when using np.load to load a npz file.

[dyvdyvaass]

https://blog.csdn.net/qq_37278761/article/details/121493618