pytorch基础教程-模型编写（二）

对于一些重复的layer可以采用nn.Sequential先进行封装layer，然后调用_make_layer_进行实现

以ResNet18/34为例

模型图：

ResNet中将一个跨层直连的单元叫做Residual block,其结构如下，当输入和输出的通道不一致的时候，需要一个专门的单元将二者转成一致，使其可以相加。

实现：

class BasicBlock(nn.Module):
    expansion = 1 

    def __init__(self,in_channels,out_channels,stride=1,downsample=None):
        super(BasicBlock,self).__init__()
        self.conv1 = conv3x3(
            in_channels,out_channels,stride=stride
        )
        self.bn1 = nn.BatchNorm2d(out_channels)
        self.relu = nn.ReLU(inplace=True)

        self.conv2 = conv3x3(
            in_channels,out_channels
        )
        self.bn2 = nn.BatchNorm2d(in_channels)
        self.downsample = downsample
        self.stride = stride

    def forward(self,x):
        residual = x
        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            residual = self.downsample(x)
        out += residual
        out = self.relu(out)

        return out 

_make_layer实现

    def _make_layer(self,block,out_channels,blocks,stride=1):

        ## 虚线部分
        downsample = None
        if stride!=1 or self.in_channels != out_channels*block.expansion:

            downsample = nn.Sequential(
                nn.Conv2d(
                    self.in_channels,out_channels*block.expansion,kernel_size=1,
                    stride=stride,bias=False 
                ),
                nn.BatchNorm2d(out_channels*block.expansion)
            )
        
        layers = []
        layers.append(block(self.in_channels,out_channels,stride,downsample))
        self.in_channels = out_channels*block.expansion 

        for i in range(1,blocks):
            layers.append(block(self.in_channels,out_channels))
        
        return nn.Sequential(*layers)

 实现：

class ResNet(nn.Module):

    def __init__(self,block,layers,num_classes=1000):

        self.in_channels = 64
        super(ResNet,self).__init__()
        self.conv1 = nn.Conv2d(
            3,64,kernel_size=7,stride=2,padding=3,bias=False
        )
        self.bn1 = nn.BatchNorm2d(64)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3,stride=2,padding=1)
        self.layer1 = self._make_layer(block,64,layers[0])
        self.layer2 = self._make_layer(block,128,layers[1],stride=2)
        self.layer3 = self._make_layer(block,256,layers[2],stride=2)
        self.layer4 = self._make_layer(block,512,layers[3],stride=2)
        self.avgpool = nn.AvgPool2d(7,stride=1)
        self.fc = nn.Linear(512*block.expansion,num_classes)


    def _make_layer(self,block,out_channels,blocks,stride=1):

        
        downsample = None
        if stride!=1 or self.in_channels != out_channels*block.expansion:

            downsample = nn.Sequential(
                nn.Conv2d(
                    self.in_channels,out_channels*block.expansion,kernel_size=1,
                    stride=stride,bias=False 
                ),
                nn.BatchNorm2d(out_channels*block.expansion)
            )
        
        layers = []
        ## 虚线部分
        layers.append(block(self.in_channels,out_channels,stride,downsample))
        ## 更新 in_channels
        self.in_channels = out_channels*block.expansion 

        for i in range(1,blocks):
            layers.append(block(self.in_channels,out_channels))
        
        return nn.Sequential(*layers)
    

    def forward(self,x):

        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)

        x2 = self.layer1(x)
        x3 = self.layer2(x2)
        x4 = self.layer3(x3)

        x5 = self.layer4(x4)
        x6 = self.avgpool(x5)
        x7 = x6.view(x6.size(0),-1)
        return self.fc(x7)

resnet18实现

def resnet18():

    model = ResNet(BasicBlock,[2,2,2,2])
    return model

resnet34实现 [3, 4, 6, 3]，把数组替换下