nn.Sequential()实例化模型的三种方法

1.直接进行实例化的问题——每层的模型没有名称，只能通过索引和下标进行访问

import torch.nn as nn

model = nn.Sequential(
    nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, padding=1),
    nn.BatchNorm2d(16),
    nn.ReLU(inplace=True),
    nn.MaxPool2d(2, 2)
)
print(model)
print(model[0])
'''运行结果为
Sequential(
  (0): Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
  (1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
  (2): ReLU(inplace=True)
  (3): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))

'''

2.给每层增加一个名称——OrderedDict有序字典

这种添加方式，虽然加入了名称，但是仍然只支持下标索引访问

model[0] 是正确的

model["conv1"] 是错误的

model = nn.Sequential(
    OrderedDict([("conv1", nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, padding=1)),
                 ("bn1", nn.BatchNorm2d(16)),
                 ("relu1", nn.ReLU(inplace=True)),
                 ("maxpool", nn.MaxPool2d(2, 2))])
)
print(model)
print(model[0])
"""
Sequential(
  (conv1): Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
  (bn1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
  (relu1): ReLU(inplace=True)
  (maxpool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
"""

3.逐层添加

model=nn.Sequential()
model.add_module("conv1",nn. Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)))
model.add_module("bn1",nn.BatchNorm2d(16))
model.add_module("relu1",nn.ReLU(inplace=True))
model.add_module("maxpool",nn.MaxPool2d(2, 2))
print(model)
print(model[0])