0

我正在尝试将 ExplainerDashboard 用于我的 PyTorch 神经网络分类器。根据GitHub的要求，我将其作为 Skorch 模型传递，如下所示，它在拟合后正确生成模型。

def get_skorch_classifier():
    X_train_m = X_train.astype(np.float32)
    y_train_m = y_train.astype(np.float32)

    X_train_df = pd.DataFrame(X_train_m, columns=X.columns)


    class MyModule(nn.Module):
      def __init__(self):
          super(MyModule, self).__init__()

          self.layer_1 = nn.Linear(298, 60)
          self.layer_2 = nn.Linear(60, 60)
          self.layer_3 = nn.Linear(60, 60)
          self.layer_4 = nn.Linear(60, 60)
          self.layer_out = nn.Linear(60, 1)

          self.relu = nn.ReLU()
          self.dropout = nn.Dropout(0.1)
          self.batchnorm1 = nn.BatchNorm1d(60, momentum=0.2)
          self.batchnorm2 = nn.BatchNorm1d(60, momentum=0.2)
          self.batchnorm3 = nn.BatchNorm1d(60, momentum=0.2)
          self.batchnorm4 = nn.BatchNorm1d(60, momentum=0.2)
          self.sigmoid = nn.Sigmoid()

      def forward(self, inputs):
          x = self.relu(self.layer_1(inputs))
          x = self.batchnorm1(x)
          x = self.dropout(x)
          x = self.relu(self.layer_2(x))
          x = self.batchnorm2(x)
          x = self.dropout(x)
          x = self.relu(self.layer_3(x))
          x = self.batchnorm3(x)
          x = self.dropout(x)
          x = self.relu(self.layer_4(x))
          x = self.batchnorm4(x)
          #x = self.dropout(x)
          x = self.layer_out(x)
          #x = self.sigmoid(x)
          return x

    model = NeuralNetBinaryClassifier(MyModule, max_epochs=10, lr=0.01, optimizer=optim.Adam)

    model.fit(X_train_m, torch.FloatTensor(y_train_m))
    return model,  X_train_df, y_train_m

model, Xm_df, ym = get_skorch_classifier()

然后，当试图将它传递给分类器解释器函数时，它会导致 TypeError

explainer = ClassifierExplainer(model, Xm_df, y_test)
ExplainerDashboard(explainer, mode='inline').run(port=8051)

> TypeError                                 Traceback (most recent call
> last) <ipython-input-17-76facb1989af> in <module>()
> ----> 1 explainer = ClassifierExplainer(model, Xm_df, y_test)
>       2 ExplainerDashboard(explainer, mode='inline').run(port=8051)
> 
> 10 frames
> /usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py in
> _call_impl(self, *input, **kwargs)    1049         if not (self._backward_hooks or self._forward_hooks or
> self._forward_pre_hooks or _global_backward_hooks    1050             
> or _global_forward_hooks or _global_forward_pre_hooks):
> -> 1051             return forward_call(*input, **kwargs)    1052         # Do not call functions when jit is used    1053         full_backward_hooks, non_full_backward_hooks = [], []
> 
> TypeError: forward() got an unexpected keyword argument 'temp'

'temp' 是我的 DataFrame 中的列名：

我的数据框的一部分

任何想法为什么会发生此错误以及如何解决它？

4

0 回答 0