Pyrotch NLLloss中的“预期2个或更多维度（得到1）”，即使张量的维度正确

Question

我正在尝试创建一个简单的神经网络，它有两个输入和两个输出，基于我创建的数学函数

我的问题是我无法计算损失

X_s是一个（600,2）数组，F是由大小为（600,2）的函数（代码未显示）生成的数组。模型如下：

class ReactorNet(nn.Module):

  def __init__(self, input_size, hidden1_size, hidden2_size, num_classes):

      super(ReactorNet, self).__init__()
      self.fc1 = nn.Linear(input_size, hidden1_size)
      self.relu1 = nn.ReLU()
      self.fc2 = nn.Linear(hidden1_size, hidden2_size)
      self.relu2 = nn.ReLU()
      self.fc3 = nn.Linear(hidden2_size, num_classes)

  def forward(self, x):
      out = self.fc1(x)
      
      out = self.relu1(out)
      out = self.fc2(out)
      out = self.relu2(out)
      out = self.fc3(out)
      return out

model = ReactorNet(2, 10, 4, 2 )
print(model)

X = torch.rand((1,2))
X = X.view(-1, 2)
output = model(X)

from torch.utils.data import TensorDataset, DataLoader
import torch.optim as optim
import torch.nn.functional as func

tensor_x = torch.Tensor(X_s) # transform to torch tensor
tensor_y = torch.Tensor(F)

my_dataset = TensorDataset(tensor_x,tensor_y) # create your datset
print(my_dataset)
train_set, val_set = torch.utils.data.random_split(my_dataset, [550, 50])

learn_rate = optim.Adam(model.parameters(), lr=0.001)
epochs = 1

for i in range(epochs):
    for data in train_set:
        input, output = data
        print(input)
        print(output)   
        model.zero_grad()
        result = model(input.view(-1,2).data)
        result = result.data
        result = result.squeeze(0)
        print(result)
        loss = func.nll_loss(output, result)
        loss.backward()
        learn_rate.step()
    print(loss)


# Test the network
model.eval()

产生的错误发生在NLL损失线上： Expected 2 or more dimensions (got 1)。打印这些张量表明它们似乎具有正确的维度

tensor([0.7197, 0.0468])
tensor([0.3284, 0.2165])
tensor([-0.0252, -0.2400])