PyTorch 之 class torch.nn.Sequential(* args)

nn.Sequential()对象是类似keras的前馈模型的对象，可以为之添加层实现前馈神经网络。可将其看作是一个时序容器。Modules 会以他们传入的顺序被添加到容器中。当然，也可以传入一个OrderedDict。

模型建立有三种方法：

① nn.Sequential()：对象.add_module(层名，层class的实例）

net1 = nn.Sequential()
net1.add_module('conv1', nn.Conv2d(1, 3, 5))
net1.add_module('batchnorm', nn.BatchNorm2d(3))
net1.add_module('relu1', nn.ReLU())

② nn.Sequential (*n个层class的实例)

net2 = nn.Sequential(
        nn.Conv2d(1, 3, 5),
        nn.BatchNorm2d(3),
        nn.ReLU()
        )

③ nn.Sequential (OrderedDict([*n个(层名，层class的实例)]))

from collections import OrderedDict

net3 = nn.Sequential(OrderedDict([
          ('conv1', nn.Conv2d(1, 3, 5)),
          ('batchnorm', nn.BatchNorm2d(3)),
          ('relu1', nn.ReLU())
        ]))

想查看模型具体情况，print 对象即可：

print('net1:', net1)
print('net2:', net2)
print('net3:', net3)

net1: Sequential(
  (conv1): Conv2d (1, 3, kernel_size=(5, 5), stride=(1, 1))
  (batchnorm): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True)
  (relu1): ReLU()
)

net2: Sequential(
  (0): Conv2d (1, 3, kernel_size=(5, 5), stride=(1, 1))
  (1): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True)
  (2): ReLU()
)

net3: Sequential(
  (conv1): Conv2d (1, 3, kernel_size=(5, 5), stride=(1, 1))
  (batchnorm): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True)
  (relu1): ReLU()
)