Hi, I am new to pytorch and trying to make CNN model.
I got a help by chatGPT, so I didn’t sincerely understand the whole code.
I got the ‘Expected input batch_size () to match target batch_size ()’ error until yesterday, and now I get the ‘mat1 and mat2 shapes cannot be multiplied (50176x128 and 8192x512)’ error.
I didn’t change the code today, so I don’t know what make this problem.
Also, the batch size error hadn’t been solved for a week. I tried all of the solution I had found, but they didn’t work at all.
Below is my whole code.
class CNN(nn.Module):
def __init__(self):
super(CNN, self).__init__()
self.conv1 = nn.Conv2d(in_channels=3, out_channels=32, kernel_size=3, stride=1, padding=1)
self.conv2 = nn.Conv2d(in_channels=32, out_channels=64, kernel_size=3, stride=1, padding=1)
self.conv3 = nn.Conv2d(in_channels=64, out_channels=128, kernel_size=3, stride=1, padding=1)
self.pool = nn.MaxPool2d(kernel_size=2, stride=2)
self.fc1 = nn.Linear(in_features=128*8*8, out_features=512)
self.fc2 = nn.Linear(in_features=512, out_features=10)
self.relu = nn.ReLU()
def forward(self, x):
x = self.conv1(x)
x = self.relu(x)
x = self.pool(x)
x = self.conv2(x)
x = self.relu(x)
x = self.pool(x)
x = self.conv3(x)
x = self.relu(x)
x = self.pool(x)
x = x.view(-1, 128)
x = self.fc1(x)
x = self.relu(x)
x = self.fc2(x)
return x
batch_size = 64
num_workers = 2
train_size = batch_size * (len(dataset) // batch_size)
test_size = len(dataset) - train_size
train_dataset, test_dataset = random_split(dataset, [train_size, test_size])
train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True, pin_memory=True, num_workers=num_workers)
test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False, pin_memory=True, num_workers=num_workers)
model = CNN()
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=lr, momentum=momentum)
lr = 0.001
momentum = 0.9
num_epochs = 10
device = 'cuda' if torch.cuda.is_available() else 'cpu'
model.to(device)
for epoch in range(num_epochs):
model.train()
train_loss = 0.0
train_correct = 0
total = 0
for batch_idx, (inputs, targets) in enumerate(train_loader):
inputs, targets = inputs.to(device), targets.to(device)
optimizer.zero_grad()
outputs = model(inputs)
loss = criterion(outputs, targets)
loss.backward()
optimizer.step()
train_loss += loss.item() * targets.size(0)
_, predicted = outputs.max(1)
total += targets.size(0)
train_correct += predicted.eq(targets).sum().item()
train_loss = train_loss / train_size
train_acc = train_correct / total
model.eval()
test_loss = 0.0
test_correct = 0
total = 0
with torch.no_grad():
for batch_idx, (inputs, targets) in enumerate(test_loader):
inputs, targets = inputs.to(device), targets.to(device)
outputs = model(inputs)
loss = criterion(outputs, targets)
test_loss += loss.item() * targets.size(0)
_, predicted = outputs.max(1)
total += targets.size(0)
test_correct += predicted.eq(targets).sum().item()
test_loss = test_loss / test_size
test_acc = test_correct / total
print('Epoch [{}/{}], Train Loss: {:.4f}, Train Acc: {:.4f}, Test Loss: {:.4f}, Test Acc: {:.4f}'.format(
epoch+1, num_epochs, train_loss, train_acc, test_loss, test_acc))