Question

我在下面发布的代码只是应用程序的一小部分：

def train(self, training_reviews, training_labels):
        
        # make sure out we have a matching number of reviews and labels
        assert(len(training_reviews) == len(training_labels))
        
        # Keep track of correct predictions to display accuracy during training 
        correct_so_far = 0
        
        # Remember when we started for printing time statistics
        start = time.time()
        
        
        criterion = nn.CrossEntropyLoss()
        optimizer =  torch.optim.SGD(self.parameters(), lr=self.learning_rate)

        # loop through all the given reviews and run a forward and backward pass,
        # updating weights for every item
        for i in range(len(training_reviews)):
            
            # TODO: Get the next review and its correct label
            review = training_reviews[i]
            label = training_labels[i]
            print('processing item ',i)
            self.update_input_layer(review)
            output = self.forward(torch.from_numpy(self.layer_0).float()) 
            target = self.get_target_for_label(label)
            print('output ',output)
            print('target ',target)
            loss = criterion(output, target)

...
mlp = SentimentNetwork(reviews[:-1000],labels[:-1000], learning_rate=0.1)
mlp.train(reviews[:-1000],labels[:-1000])

，并且在评估时以标题行中的异常结尾：

loss = criterion(output, target)

在此之前，变量如下：

output  tensor([[0.5803]], grad_fn=<SigmoidBackward>)
target  1

Answer 1

目标应为torch.Tensor变量。使用torch.tensor([target])。

此外，您可能希望使用批次（因此，有N个样本，torch.tensor的形状为(N,)，与target相同）。

有关PyTorch的信息，请参见introductory tutorial，因为您可能没有使用批处理，未运行优化器或未使用torch.utils.data.Dataset和torch.utils.data.DataLoader。

pytorch训练循环以“ int”对象结束，没有属性“ size”异常

1 个答案: