“MaskedLMOutput”物件沒有屬性“視圖”

我寫了這個：

def forward(self, x):
    x = self.bert(x)
    
    x = x.view(x.shape[0], -1)
    x = self.fc(self.dropout(self.bn(x)))
    return x

但它不能很好地作業，錯誤是“MaskedLMOutput”物件沒有屬性“視圖”。我正在考慮輸入可能不是“張量”型別，所以我將其更改如下：

def forward(self, x):
        x = torch.tensor(x)     # this part
        x = self.bert(x)
        
        x = x.view(x.shape[0], -1)
        x = self.fc(self.dropout(self.bn(x)))
        return x

但它仍然出錯，同樣的錯誤'MaskedLMOutput'物件沒有屬性'view'。

有人可以告訴我如何解決這個問題嗎？非常感謝。

整個錯誤資訊在這里：

 ------------------------------------------------------------------------
    AttributeError                            Traceback (most recent call last)
    Input In [5], in <cell line: 8>()
          6 optimizer = optim.Adam(bert_punc.parameters(), lr=learning_rate_top)
          7 criterion = nn.CrossEntropyLoss()
    ----> 8 bert_punc, optimizer, best_val_loss = train(bert_punc, optimizer, criterion, epochs_top, 
          9     data_loader_train, data_loader_valid, save_path, punctuation_enc, iterations_top, best_val_loss=1e9)
    
    Input In [3], in train(model, optimizer, criterion, epochs, data_loader_train, data_loader_valid, save_path, punctuation_enc, iterations, best_val_loss)
         17 inputs.requires_grad = False
         18 labels.requires_grad = False
    ---> 19 output = model(inputs)
         20 loss = criterion(output, labels)
         21 loss.backward()
    
    File ~\anaconda3\lib\site-packages\torch\nn\modules\module.py:1110, in Module._call_impl(self, *input, **kwargs)
       1106 # If we don't have any hooks, we want to skip the rest of the logic in
       1107 # this function, and just call forward.
       1108 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
       1109         or _global_forward_hooks or _global_forward_pre_hooks):
    -> 1110     return forward_call(*input, **kwargs)
       1111 # Do not call functions when jit is used
       1112 full_backward_hooks, non_full_backward_hooks = [], []
    
    File ~\anaconda3\lib\site-packages\torch\nn\parallel\data_parallel.py:166, in DataParallel.forward(self, *inputs, **kwargs)
        163     kwargs = ({},)
        165 if len(self.device_ids) == 1:
    --> 166     return self.module(*inputs[0], **kwargs[0])
        167 replicas = self.replicate(self.module, self.device_ids[:len(inputs)])
        168 outputs = self.parallel_apply(replicas, inputs, kwargs)
    
    File ~\anaconda3\lib\site-packages\torch\nn\modules\module.py:1110, in Module._call_impl(self, *input, **kwargs)
       1106 # If we don't have any hooks, we want to skip the rest of the logic in
       1107 # this function, and just call forward.
       1108 if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
       1109         or _global_forward_hooks or _global_forward_pre_hooks):
    -> 1110     return forward_call(*input, **kwargs)
       1111 # Do not call functions when jit is used
       1112 full_backward_hooks, non_full_backward_hooks = [], []
    
    File D:\BertPunc-original\model.py:21, in BertPunc.forward(self, x)
         18 x = torch.tensor(x)
         19 x = self.bert(x)
    ---> 21 x = x.view(x.shape[0], -1)
         22 x = self.fc(self.dropout(self.bn(x)))
         23 return x
    
    AttributeError: 'MaskedLMOutput' object has no attribute 'view'

uj5u.com熱心網友回復：

我認為這應該可以幫助您解決錯誤。https://stackoverflow.com/a/72601533/13748930 self.bert(x) 之后的輸出是 MaskedLMOutput 類的一個物件。

標籤：Python 张量流 火炬 拥抱变形金刚 伯特语言模型

上一篇：當我從D類呼叫C().__init__()時，為什么C().__init__()的結果會列印兩次？

下一篇：jinja2.exceptions.UndefinedError:'forms.ContactFormobject'沒有屬性'message'