Extract features by small version of Resnet18

I want to extract features after using a small version of ResNet-18 pretrained on ImageNet. My code is below:
import torchvision.models as model
model = model.resnet18(pretrained = True)
model = nn.Sequential(*list(model.children())[:-1])

for input, target in train_loader:
____with torch.no_grad():
________output = model(input)
________output = output.view(output.size(0),-1)

However, it seems do not match with the statement: “we use a smaller version of ResNet18, with three times less feature maps across all layers”
Please guide me how can I use “a smaller version of Resnet18”?