python - pytorch "log_softmax_lastdim_kernel_impl"没有为 'torch.LongTensor' 实现

Question

我正在尝试使用自己的数据集根据 https://github.com/bentrevett/pytorch-sentiment-analysis/blob/master/5%20-%20Multi-class%20Sentiment%20Analysis.ipynb 对文本进行分类.我的数据集是句子的 csv 和与之关联的类。有 6 个不同的类:

sent                      class
'the fox is brown'        animal
'the house is big'        object
'one water is drinkable'  water
...

运行时:

N_EPOCHS = 5

best_valid_loss = float('inf')

for epoch in range(N_EPOCHS):

    start_time = time.time()
    print(start_time)
    train_loss, train_acc = train(model, train_iterator, optimizer, criterion)
    print(train_loss.type())
    print(train_acc.type())
    valid_loss, valid_acc = evaluate(model, valid_iterator, criterion)

    end_time = time.time()

    epoch_mins, epoch_secs = epoch_time(start_time, end_time)

    if valid_loss < best_valid_loss:
        best_valid_loss = valid_loss
        torch.save(model.state_dict(), 'tut5-model.pt')

    print(f'Epoch: {epoch+1:02} | Epoch Time: {epoch_mins}m {epoch_secs}s')
    print(f'\tTrain Loss: {train_loss:.3f} | Train Acc: {train_acc*100:.2f}%')
    print(f'\t Val. Loss: {valid_loss:.3f} |  Val. Acc: {valid_acc*100:.2f}%')

，我收到以下错误

RuntimeError: "log_softmax_lastdim_kernel_impl" not implemented for 'torch.LongTensor'

指向:

<ipython-input-38-9c6cff70d2aa> in train(model, iterator, optimizer, criterion)
     14         print('pred'+ predictions.type())
     15         #batch.label = batch.label.type(torch.LongTensor)
---> 16         loss = criterion(predictions.long(), batch.label)**

此处发布的解决方案 https://github.com/pytorch/pytorch/issues/14224建议我需要使用 long/int。

我必须在 ** 行添加 .long() 以修复此早期错误:

RuntimeError:预期的标量类型 Long 对象，但参数 #2 'target' 得到标量类型 Float

具体的代码行是:

  def train(model, iterator, optimizer, criterion):
    epoch_loss = 0
    epoch_acc = 0

    model.train()

    for batch in iterator:

        optimizer.zero_grad()

        predictions = model(batch.text)
        print('pred'+ predictions.type())
        #batch.label = batch.label.type(torch.LongTensor)
        loss = criterion(predictions.long(), batch.label)**

        acc = categorical_accuracy(predictions, batch.label)

        loss.backward()

        optimizer.step()

        epoch_loss += loss.item()
        epoch_acc += acc.item()

    return epoch_loss / len(iterator), epoch_acc / len(iterator)

请注意，** 最初是 loss = criterion(predictions, batch.label)

还有其他解决此问题的建议吗？

Answer 1

criterion 在您的 notebook 中定义为 torch.nn.CrossEntropyLoss() .如 CrossEntropyLoss 的文档中所述，它期望模型为每个“K”类返回概率值，并将地面实况标签的相应值作为输入。现在，概率值是浮点张量，而真实标签应该是代表类的长张量(类不能是 float ，例如 2.3 不能代表类)。因此:

loss = criterion(predictions, batch.label.long())

应该可以。