当前位置：首页 > article >正文

【Pytorch报错】AttributeError: cannot assign module before Module.init() call

article 2025/1/8 22:20:56

代码：

import torch.nn as nn

class Model(nn.Module):
    def __init__(self, input_dim, output_dim):
        self.linear = nn.Linear(input_dim, output_dim) 
        
    def forward(self, x):
        out = self.linear(x)
        return out

model = Model(1, 1)

报错：

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[20], line 11
      8         out = self.linear(x)
      9         return out
---> 11 model = Model(1, 1)

Cell In[20], line 5, in Model.__init__(self, input_dim, output_dim)
      4 def __init__(self, input_dim, output_dim):
----> 5     self.linear = nn.Linear(input_dim, output_dim)

File C:\PyCode\venv_ai\lib\site-packages\torch\nn\modules\module.py:1214, in Module.__setattr__(self, name, value)
   1212 if isinstance(value, Module):
   1213     if modules is None:
-> 1214         raise AttributeError(
   1215             "cannot assign module before Module.__init__() call")
   1216     remove_from(self.__dict__, self._parameters, self._buffers, self._non_persistent_buffers_set)
   1217     modules[name] = value

AttributeError: cannot assign module before Module.__init__() call

原因：
这个错误信息表明在Model类的构造函数中，你试图在调用父类nn.Module的构造函数之前就赋值一个模块（nn.Linear）。在PyTorch中，你应该首先调用父类的构造函数，然后再定义子类的属性。

修改：

import torch.nn as nn

class Model(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()  # 调用父类的构造函数
        self.linear = nn.Linear(input_dim, output_dim)  # 然后定义子类的属性
        
    def forward(self, x):
        out = self.linear(x)
        return out

model = Model(1, 1)