0

您好，我在尝试使用 pytorch 训练 RNN 分类器时遇到此错误。我希望有人可以帮助我解决它或解释我的原因。

train_loader2=embeddings_train

train_loader2=torch.from_numpy(embeddings_train)
train_loader2 = DataLoader(dataset=train_dataset, batch_size=batch_size, shuffle=True)
y_train=np.array(y_train)
targets2= torch.from_numpy(y_train)

# Initialize network
model = BRNN(input_size, hidden_size, num_layers, num_classes).to(device)

# Loss and optimizer
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=learning_rate)

# Train Network
for epoch in range(num_epochs):
    for i, (dataf, targets2) in enumerate(train_loader2):
        # Get data to cuda if possible
        dataf = dataf.to(device=device).squeeze(1)
        targets2 = targets2.to(device=device)

        # forward
        scores = model(dataf)
        loss = criterion(scores, targets2)

        # backward
        optimizer.zero_grad()
        loss.backward()

        # gradient descent or adam step
        optimizer.step()

这是它发生的错误

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-42-d435f4563f2e> in <module>
     15 # Train Network
     16 for epoch in range(num_epochs):
---> 17     for i, (dataf, targets2) in enumerate(train_loader2):
     18         # Get data to cuda if possible
     19         dataf = dataf.to(device=device).squeeze(1)

ValueError: too many values to unpack (expected 2)
4

0 回答 0