No training

[jihyeheo]

Hello, i'm jihyeheo.
I am trying to do classification using densenet.
However, for some reason, it doesn't seem to be learning. yields the same result.

Why?
This is my code

count = 0
num_epochs = 100
loss_list = []
iteration_list = []
accuracy_list = []
train_loss = []
for epoch in range(num_epochs):
    train_losses = 0.0
    for i, (images, labels) in enumerate(train_loader):
        model.train()
        train = Variable(images.view(-1,1,53,53,53))
        labels = Variable(labels.type(torch.FloatTensor))
        optimizer.zero_grad()
        outputs = model(train.type(torch.FloatTensor)).view(-1,1)

        print(outputs.)

        loss = error(outputs, labels)
        train_losses += loss
       
        loss.backward()
        optimizer.step()

        # Calculate Accuracy         
        correct = 0
        total = 0
        # Iterate through test dataset
        for images, labels in test_loader:
            model.eval()
            test = Variable(images.view(-1,1,53,53,53))
            # Forward propagation
            outputs = model(test.type(torch.FloatTensor)).view(-1,1)
            predicted = (outputs.data > 0.5)

            #print(predicted)

            # Total number of labels
            total += len(labels)
            correct += (predicted == labels).sum()

        accuracy = 100 * correct / float(total)
        #print(loss.data)
        #print(accuracy)

        # store loss and iteration
        loss_list.append(loss.item())
        iteration_list.append(count)
        accuracy_list.append(accuracy.item())
   
        # Print Loss
    print('Loss: {}  Accuracy: {} %'.format(loss.data, accuracy))
    print(train_losses)
    train_loss.append(train_losses.item())

[KumoLiu]

Hi @jihyeheo, if you are new to classification, would suggest you follow these tutorials(3d classification and 2d classification) as a start. Hope it can help you, thanks!