Accuracy value more than 1 with nn.BCEWithLogitsLoss() loss function pytorch in Binary Classifier

Question

I am trying to use nn.BCEWithLogitsLoss() for model which initially used nn.CrossEntropyLoss(). However, after doing some changes to the training function to accommodate the nn.BCEWithLogitsLoss() loss function the model accuracy values are shown as more than 1. Please find the code below.

# Data augmentation and normalization for training
# Just normalization for validation
data_transforms = {
    'train': transforms.Compose([
        transforms.RandomResizedCrop(224),
        transforms.RandomHorizontalFlip(),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    ]),
    'val': transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    ]),
}

data_dir = '/kaggle/input/catsndogsorg/hymenoptera_data'
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),
                                          data_transforms[x])
                  for x in ['train', 'val']}
dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,
                                             shuffle=True, num_workers=4)
              for x in ['train', 'val']}
dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}
class_names = image_datasets['train'].classes

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

#############

def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
    since = time.time()

    best_model_wts = copy.deepcopy(model.state_dict())
    best_acc = 0.0

    for epoch in range(num_epochs):
        print(f'Epoch {epoch}/{num_epochs - 1}')
        print('-' * 10)

        # Each epoch has a training and validation phase
        for phase in ['train', 'val']:
            if phase == 'train':
                model.train()  # Set model to training mode
            else:
                model.eval()   # Set model to evaluate mode

            running_loss = 0.0
            running_corrects = 0

            # Iterate over data.
            for inputs, labels in dataloaders[phase]:
                inputs = inputs.to(device)
                labels = labels.to(device).unsqueeze(1)

                # zero the parameter gradients
                optimizer.zero_grad()

                # forward
                # track history if only in train
                with torch.set_grad_enabled(phase == 'train'):
                    outputs = model(inputs)
                    _, preds = torch.max(outputs, 1)
                    #print(outputs, labels)
                    loss = criterion(outputs, labels.float())
                    print(loss)
                    # backward + optimize only if in training phase
                    if phase == 'train':
                        loss.backward()
                        optimizer.step()

                # statistics
                running_loss += loss.item() * inputs.size(0)
                running_corrects += torch.sum(preds == labels.data)
            if phase == 'train':
                scheduler.step()

            epoch_loss = running_loss / dataset_sizes[phase]
            epoch_acc = running_corrects.double() / dataset_sizes[phase]

            print(f'{phase} Loss: {epoch_loss:.4f} Acc: {epoch_acc:.4f}')

            # deep copy the model
            if phase == 'val' and epoch_acc > best_acc:
                best_acc = epoch_acc
                best_model_wts = copy.deepcopy(model.state_dict())

        print()

    time_elapsed = time.time() - since
    print(f'Training complete in {time_elapsed // 60:.0f}m {time_elapsed % 60:.0f}s')
    print(f'Best val Acc: {best_acc:4f}')

    # load best model weights
    model.load_state_dict(best_model_wts)
    return model

EDIT:Model pipeline

model_ft = models.resnet18(weights='ResNet18_Weights.DEFAULT')
num_ftrs = model_ft.fc.in_features
# Here the size of each output sample is set to 2.
# Alternatively, it can be generalized to nn.Linear(num_ftrs, len(class_names)).
model_ft.fc = nn.Linear(num_ftrs, 1)

model_ft = model_ft.to(device)

criterion = nn.BCEWithLogitsLoss()

# Observe that all parameters are being optimized
optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)

# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)

model_ft = train_model(model_ft, criterion, optimizer_ft, 
     exp_lr_scheduler,num_epochs=25)

The outputs of training loop:

outputs shape:  torch.Size([4, 1])
labels shape:  torch.Size([4, 1])

logits: tensor(0.3511,grad_fn<BinaryCrossEntropyWithLogitsBackward0>)

train Loss: 1.0000 Acc: 2.0164
val Loss: 1.0000 Acc: 1.8105

Would anyone be able to help me in this matter please.

Thanks & Best Regards AMJS

Answer 1

I didn't understand why you are using torch.max as you have one output. Anyway, you should use squeeze before comparing, so this line:

running_corrects += torch.sum(preds == labels.data)

should become

 running_corrects += torch.sum(preds == labels.squeeze())

to see why:

labels = torch.tensor([[0],
        [0],
        [0],
        [1]])

preds = torch.tensor([0, 0, 0, 0])
print(torch.sum(preds == labels)) # output 12
print(torch.sum(preds == labels.squeeze())) # output 3