DefaultPCUAllocator: not enough memory

Question

I want to do some CNN with Pytorch, but I got this error:

RuntimeError: [enforce fail at C:\cb\pytorch_1000000000000\work\c10\core\impl\alloc_cpu.cpp:81] data. DefaultCPUAllocator: not enough memory: you tried to allocate 412876800 bytes.

The dataset I am using is 3410 images consists of 0-9, A-Z, and a-z. I assume the dataset is not that large to the point that I don't have enough RAM to process them.

I read about some possible solutions to reduce batch size, but when I reduce the batch_size to 16 and run the file, nothing happens. Does anyone have any idea how to solve this?

Here is my code:

class WritingDataset(Dataset):
    def __init__(self, csv_file, root_dir, transform=None):
        self.annotations = pd.read_csv(csv_file)
        self.root_dir = root_dir
        self.transform = transform
        
    def __len__(self):
        return len(self.annotations)

    def __getitem__(self, index):
        img_path = os.path.join(self.root_dir, self.annotations.iloc[index, 0])
        image = io.imread(img_path)
        y_label = torch.tensor(int(self.annotations.iloc[index, 1]))

        if self.transform:
            image = self.transform(image)

        return (image, y_label)


# Set device
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Hyperparameters
in_channel = 3
num_classes = 2
learning_rate = 1e-3
batch_size = 32
num_epochs = 5

# Load Data
dataset = WritingDataset(
    csv_file='english.csv',
    root_dir='Img',
    transform=transforms.ToTensor()
)

train_set, test_set = torch.utils.data.random_split(dataset, [3000, 410])
train_loader = DataLoader(dataset=train_set, batch_size=batch_size, shuffle=True)
test_loader = DataLoader(dataset=test_set, batch_size=batch_size, shuffle=True)

# Model
model = torchvision.models.googlenet(pretrained=True)
model.to(device)

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

# Train Network
for epoch in range(num_epochs):
    losses = []

    for batch_idx, (data, targets) in enumerate(train_loader):
        # Get data to cuda if possible
        data = data.to(device=device)
        targets = targets.to(device=device)

        # forward
        scores = model(data)
        loss = criterion(scores, targets)

        losses.append(loss.item())

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

        # gradient descent or adam step
        optimizer.step()

    print(f"Cost at epoch {epoch} is {sum(losses)/len(losses)}")

# Check accuracy on training & test
def check_accuracy(loader, model):
    num_correct = 0
    num_samples = 0
    model.eval()

    with torch.no_grad():
        for x, y in loader:
            x = x.to(device=device)
            y = y.to(device=device)

            scores = model(x)
            _, predictions = scores.max(1)
            num_correct += (predictions == y).sum()
            num_samples += predictions.size(0)

        print(
            f"Got {num_correct} / {num_samples} with accuracy {float(num_correct)/float(num_samples)*100:.2f}"
        )

    model.train()

Answer 1

If you have enough VRAM memory, try to increase the pagefile (virtual memory) size. The better option is to set 'System managed', or manually increase it to higher values.