I want to extract some layers from a model, so I write
nn.Sequential(list(model.children())[:7])
but get error: list is not a Moudule subclass. I know I should write
nn.Sequential(*list(model.children)[:7])
but why I must add the * ??? If I just wanna to get the [] including the layers, I must write
layer = list(model.children())[:7]
but not
layer = *list(model.children())[:7]
In this situation, the * does not work and get error
layer = *list(model_ft.children())[:3]
^
SyntaxError: can't use starred expression here
Why???
list(model.children())[:7] returns a list, but the input of nn.Sequential() requires the modules to be an OrderedDict or to be added directly, not in a python list.
nn.SequentialModules will be added to it in the order they are passed in the constructor. Alternatively, an OrderedDict of modules can be passed in.
# nn.Sequential(list(model.children)[:3]) means, which is wrong
nn.Sequential([Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3)),
ReLU(inplace=True),
MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)])
# nn.Sequential(*list(model.children)[:3]) mean
nn.Sequential(Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3)),
ReLU(inplace=True),
MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False))
That's why you need to unpack the list using *. It can only be used inside a function, hence, it doesn't work in your last case. Read * in python