PyTorch的Modules的使用

传送门

模型的参数!

model.parameters() / model.named_parameters()

for parameter in m.named_parameters():
  print(parameter)
: ('weight', Parameter containing:
tensor([[ 1.0597,  1.1796,  0.8247],
        [-0.5080, -1.2635, -1.1045],
        [ 0.0593,  0.2469, -1.4299],
        [-0.4926, -0.5457,  0.4793]], requires_grad=True))
('bias', Parameter containing:
tensor([ 0.3634,  0.2015, -0.8525], requires_grad=True))

Building Blocks

net = nn.Sequential(
  MyLinear(4, 3),
  nn.ReLU(),
  MyLinear(3, 1)
)

sample_input = torch.randn(4)
net(sample_input)
: tensor([-0.6749], grad_fn=<AddBackward0>)

children() or named_children()

中间层（直接层）

net = Net()
for child in net.named_children():
  print(child)
: ('l0', MyLinear())
('l1', MyLinear())

modules() and named_modules()

递归地表示出所有的中间层（以及中间层的子层）

class BigNet(nn.Module):
  def __init__(self):
    super().__init__()
    self.l1 = MyLinear(5, 4)
    self.net = Net()
  def forward(self, x):
    return self.net(self.l1(x))

big_net = BigNet()
for module in big_net.named_modules():
  print(module)


: ('', BigNet(
  (l1): MyLinear()
  (net): Net(
    (l0): MyLinear()
    (l1): MyLinear()
  )
))
('l1', MyLinear())
('net', Net(
  (l0): MyLinear()
  (l1): MyLinear()
))
('net.l0', MyLinear())
('net.l1', MyLinear())