文章目錄
- 模型介紹
- resnet18模型流程
- 總結
- resnet50
- 總結
resnet和resnext的架構基本相同的，這裡先學習下resnet的建構，感覺高度子產品化，很友善。本文算是對
PyTorch源碼解讀之torchvision.modelsResNet代碼的詳細了解，另外，強烈推薦這位大神的PyTorch的教程！
模型介紹
resnet的模型可以直接通過torchvision導入，可以通過pretrained設定是否導入預訓練的參數。
import torchvision
model = torchvision.models.resnet50(pretrained=False) 如果選擇導入，resnet50、resnet101和resnet18等的模型函數十分簡潔并且隻有ResNet的參數不同，隻是需要導入預訓練參數時，調用
load_state_dict
加載
model_zoo.load_url
下載下傳的參數，這裡
model_urls
是一個維護不同模型參數下載下傳位址的字典。
def resnet18(pretrained=False, **kwargs):
"""Constructs a ResNet-18 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))
return model
def resnet50(pretrained=False, **kwargs):
"""Constructs a ResNet-50 model.
Args:
pretrained (bool): If True, returns a model pre-trained on ImageNet
"""
model = ResNet(Bottleneck, [3, 4, 6, 3], **kwargs)
if pretrained:
model.load_state_dict(model_zoo.load_url(model_urls['resnet50']))
return model
model_urls = {
'resnet18': 'https://download.pytorch.org/models/resnet18-5c106cde.pth',
'resnet34': 'https://download.pytorch.org/models/resnet34-333f7ec4.pth',
'resnet50': 'https://download.pytorch.org/models/resnet50-19c8e357.pth',
'resnet101': 'https://download.pytorch.org/models/resnet101-5d3b4d8f.pth',
'resnet152': 'https://download.pytorch.org/models/resnet152-b121ed2d.pth',
} 接下來我們看下重點，也就是ResNet，ResNet的組成是:基礎子產品Bottleneck/Basicblock，通過_make_layer生成四個的大的layer，然後在forward中排序。
__init__的兩個重要參數，block和layers，block有兩種(Bottleneck/Basicblock)，不同模型調用的類不同在resnet50、resnet101、resnet152中調用的是Bottleneck類，而在resnet18和resnet34中調用的是BasicBlock類，在後面我們詳細了解。layers是包含四個元素的清單，每個元素分别是_make_layer生成四個的大的layer的包含的resdual子結構的個數，在resnet50可以看到清單是 [3, 4, 6, 3]。
_make_layer包含四個參數，第一個參數是block的類型，第二個參數planes是輸出的channel數，第三個參數blocks每個blocks中包含多少個residual子結構，也就是上述清單layers所存儲的數字，第四個參數為步長。
class ResNet(nn.Module):
def __init__(self, block, layers, num_classes=1000):
self.inplanes = 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)
for m in self.modules():
if isinstance(m, nn.Conv2d):
n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
m.weight.data.normal_(0, math.sqrt(2. / n)) # 卷積參數變量初始化
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1) # BN參數初始化
m.bias.data.zero_()
def _make_layer(self, block, planes, blocks, stride=1):
downsample = None
if stride != 1 or self.inplanes != planes * block.expansion:
downsample = nn.Sequential(
nn.Conv2d(self.inplanes, planes * block.expansion,
kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(planes * block.expansion),
)
layers = []
layers.append(block(self.inplanes, planes, stride, downsample))
self.inplanes = planes * block.expansion
for i in range(1, blocks):
layers.append(block(self.inplanes, planes))
return nn.Sequential(*layers)
def forward(self, x):
x = self.conv1(x)
x = self.bn1(x)
x = self.relu(x)
x = self.maxpool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.avgpool(x)
x = x.view(x.size(0), -1)
x = self.fc(x)
return x 接下來我們看下兩種block：Bottleneck/Basicblock，他們最重要的是resdual的結構。所有的模型都繼承
torch.nn.Module
，bottleneck改寫了__init__和forward()，forward()中的
out += residual
就是element-wise add的操作。Bottleneck需要了解的有兩處：expansion=4和downsample(下采樣)。關于下采樣的理論我也不清楚，我們後面直接通過代碼來了解吧。
class Bottleneck(nn.Module):
expansion = 4
def __init__(self, inplanes, planes, stride=1, downsample=None):
super(Bottleneck, self).__init__()
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
self.bn1 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
padding=1, bias=False)
self.bn2 = nn.BatchNorm2d(planes)
self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
self.bn3 = nn.BatchNorm2d(planes * 4)
self.relu = nn.ReLU(inplace=True)
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)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
if self.downsample is not None:
residual = self.downsample(x)
out += residual
out = self.relu(out)
return out Basicblock的resdual包含兩個卷積層，第一層卷積層的kernel=3。
def conv3x3(in_planes, out_planes, stride=1):
"""3x3 convolution with padding"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, inplanes, planes, stride=1, downsample=None):
super(BasicBlock, self).__init__()
self.conv1 = conv3x3(inplanes, planes, stride)
self.bn1 = nn.BatchNorm2d(planes)
self.relu = nn.ReLU(inplace=True)
self.conv2 = conv3x3(planes, planes)
self.bn2 = nn.BatchNorm2d(planes)
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 resnet18模型流程
resnet調用的Resnet參數是
model = ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
Resnet – init()
self.layer1之前的變量初始化不難了解，
self.layer1=self._make_layer(block, 64, layers[0])
這裡block=Basicblock,layer[0]=2
執行_make_layer
downsample = None——if條件不滿足，downsample=None
下面建構blocks層Basicblock：
layers=[]——layers.append(Basicblock(64,64,1,downsample=None))
指派輸入channel self.inplanes = planesblock.expansion = 641 = 64
for循環建構剩下的blocks-1個residual,不傳downsample.
self.layer2 執行
self._make_layer(block, 128, layers[1], stride=2)
downsample=None
顯然if條件滿足 downsample=nn.Sequential(nn.Conv2d(64,128, kernel_size=1, stride=2, bias=False), nn.BatchNorm2d(128),
)
layers=[]——layers.append(Basicblock(64,128,2,downsample))
self.inplanes = 128*1=128
for循環建構剩下的blocks-1個residual，不傳dowmsample.
可以看出接下來layer3和layer4與layer2相似，最終構成resnet18.
總結
從layer2到layer4，每個layer第一個輸入會增加一倍channel，是以resdual會采用下采樣，而對于每一層而言，channel都是相同的，basicblock.expansion都為1，是以我們看不出其發揮的作用，我們将在resnet50研究下。如下圖，這裡沒找到resnet18，圖中的虛線就是downsample，其産生于channel變化的resdual。
resnet50
model = ResNet(Bottleneck, [3, 4, 6, 3], **kwargs)
，可以看出，resnet50采用Bottleneck子產品，并且每個大的layer的blocks數量也不同。
layer1=self._make_layer(Bottleneck, 64, 3)
if條件滿足，downsample = nn.Sequential(
nn.Conv2d(self.inplanes=64, 64 * 4,
kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(644),)
layers.append(Bottleneck(64,64,1,dowmsample)),bottleneck内經過三個卷積層Conv2d(64,64) Conv2d(64,64) Conv2d(64,644)保證每個block的輸出channel是planesexpansion,通過self.inplanes = planesblock.expansion指派後面block的輸入channel也是planes*expansion。
