186 lines
6.0 KiB
Python
186 lines
6.0 KiB
Python
"""resnet in pytorch
|
|
|
|
|
|
|
|
[1] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun.
|
|
|
|
Deep Residual Learning for Image Recognition
|
|
https://arxiv.org/abs/1512.03385v1
|
|
"""
|
|
|
|
import torch
|
|
import torch.nn as nn
|
|
|
|
from trainer.networks import register_model
|
|
|
|
|
|
class BasicBlock(nn.Module):
|
|
"""Basic Block for resnet 18 and resnet 34
|
|
|
|
"""
|
|
|
|
#BasicBlock and BottleNeck block
|
|
#have different output size
|
|
#we use class attribute expansion
|
|
#to distinct
|
|
expansion = 1
|
|
|
|
def __init__(self, in_channels, out_channels, stride=1):
|
|
super().__init__()
|
|
|
|
#residual function
|
|
self.residual_function = nn.Sequential(
|
|
nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=stride, padding=1, bias=False),
|
|
nn.BatchNorm2d(out_channels),
|
|
nn.ReLU(inplace=True),
|
|
nn.Conv2d(out_channels, out_channels * BasicBlock.expansion, kernel_size=3, padding=1, bias=False),
|
|
nn.BatchNorm2d(out_channels * BasicBlock.expansion)
|
|
)
|
|
|
|
#shortcut
|
|
self.shortcut = nn.Sequential()
|
|
|
|
#the shortcut output dimension is not the same with residual function
|
|
#use 1*1 convolution to match the dimension
|
|
if stride != 1 or in_channels != BasicBlock.expansion * out_channels:
|
|
self.shortcut = nn.Sequential(
|
|
nn.Conv2d(in_channels, out_channels * BasicBlock.expansion, kernel_size=1, stride=stride, bias=False),
|
|
nn.BatchNorm2d(out_channels * BasicBlock.expansion)
|
|
)
|
|
|
|
def forward(self, x):
|
|
return nn.ReLU(inplace=True)(self.residual_function(x) + self.shortcut(x))
|
|
|
|
class BottleNeck(nn.Module):
|
|
"""Residual block for resnet over 50 layers
|
|
|
|
"""
|
|
expansion = 4
|
|
def __init__(self, in_channels, out_channels, stride=1):
|
|
super().__init__()
|
|
self.residual_function = nn.Sequential(
|
|
nn.Conv2d(in_channels, out_channels, kernel_size=1, bias=False),
|
|
nn.BatchNorm2d(out_channels),
|
|
nn.ReLU(inplace=True),
|
|
nn.Conv2d(out_channels, out_channels, stride=stride, kernel_size=3, padding=1, bias=False),
|
|
nn.BatchNorm2d(out_channels),
|
|
nn.ReLU(inplace=True),
|
|
nn.Conv2d(out_channels, out_channels * BottleNeck.expansion, kernel_size=1, bias=False),
|
|
nn.BatchNorm2d(out_channels * BottleNeck.expansion),
|
|
)
|
|
|
|
self.shortcut = nn.Sequential()
|
|
|
|
if stride != 1 or in_channels != out_channels * BottleNeck.expansion:
|
|
self.shortcut = nn.Sequential(
|
|
nn.Conv2d(in_channels, out_channels * BottleNeck.expansion, stride=stride, kernel_size=1, bias=False),
|
|
nn.BatchNorm2d(out_channels * BottleNeck.expansion)
|
|
)
|
|
|
|
def forward(self, x):
|
|
return nn.ReLU(inplace=True)(self.residual_function(x) + self.shortcut(x))
|
|
|
|
|
|
class ResNetTail(nn.Module):
|
|
def __init__(self, block, num_block, num_classes=100):
|
|
super().__init__()
|
|
|
|
self.in_channels = 128
|
|
self.conv4_x = self._make_layer(block, 256, num_block[2], 2)
|
|
self.conv5_x = self._make_layer(block, 512, num_block[3], 2)
|
|
self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))
|
|
self.fc = nn.Linear(512 * block.expansion, num_classes)
|
|
|
|
def _make_layer(self, block, out_channels, num_blocks, stride):
|
|
strides = [stride] + [1] * (num_blocks - 1)
|
|
layers = []
|
|
for stride in strides:
|
|
layers.append(block(self.in_channels, out_channels, stride))
|
|
self.in_channels = out_channels * block.expansion
|
|
return nn.Sequential(*layers)
|
|
|
|
def forward(self, x):
|
|
output = self.conv4_x(x)
|
|
output = self.conv5_x(output)
|
|
output = self.avg_pool(output)
|
|
output = output.view(output.size(0), -1)
|
|
output = self.fc(output)
|
|
|
|
return output
|
|
|
|
|
|
class ResNet(nn.Module):
|
|
|
|
def __init__(self, block, num_block, num_classes=100, num_tails=20):
|
|
super().__init__()
|
|
self.in_channels = 64
|
|
self.conv1 = nn.Sequential(
|
|
nn.Conv2d(3, 64, kernel_size=3, padding=1, bias=False),
|
|
nn.BatchNorm2d(64),
|
|
nn.ReLU(inplace=True))
|
|
|
|
self.conv2_x = self._make_layer(block, 64, num_block[0], 1)
|
|
self.conv3_x = self._make_layer(block, 128, num_block[1], 2)
|
|
self.tails = nn.ModuleList([ResNetTail(block, num_block, 256) for _ in range(num_tails)])
|
|
self.selector = ResNetTail(block, num_block, num_tails)
|
|
self.final_linear = nn.Linear(256, 100)
|
|
|
|
def _make_layer(self, block, out_channels, num_blocks, stride):
|
|
strides = [stride] + [1] * (num_blocks - 1)
|
|
layers = []
|
|
for stride in strides:
|
|
layers.append(block(self.in_channels, out_channels, stride))
|
|
self.in_channels = out_channels * block.expansion
|
|
return nn.Sequential(*layers)
|
|
|
|
def forward(self, x, coarse_label):
|
|
output = self.conv1(x)
|
|
output = self.conv2_x(output)
|
|
output = self.conv3_x(output)
|
|
|
|
keys = []
|
|
for t in self.tails:
|
|
keys.append(t(output))
|
|
keys = torch.stack(keys, dim=1)
|
|
|
|
query = self.selector(output).unsqueeze(2)
|
|
attn = torch.nn.functional.softmax(query * keys, dim=1)
|
|
values = self.final_linear(attn * keys)
|
|
|
|
return values.sum(dim=1)
|
|
|
|
#bs = output.shape[0]
|
|
#return (tailouts[coarse_label] * torch.eye(n=bs, device=x.device).view(bs,bs,1)).sum(dim=1)
|
|
|
|
@register_model
|
|
def register_cifar_resnet18_branched(opt_net, opt):
|
|
""" return a ResNet 18 object
|
|
"""
|
|
return ResNet(BasicBlock, [2, 2, 2, 2])
|
|
|
|
def resnet34():
|
|
""" return a ResNet 34 object
|
|
"""
|
|
return ResNet(BasicBlock, [3, 4, 6, 3])
|
|
|
|
def resnet50():
|
|
""" return a ResNet 50 object
|
|
"""
|
|
return ResNet(BottleNeck, [3, 4, 6, 3])
|
|
|
|
def resnet101():
|
|
""" return a ResNet 101 object
|
|
"""
|
|
return ResNet(BottleNeck, [3, 4, 23, 3])
|
|
|
|
def resnet152():
|
|
""" return a ResNet 152 object
|
|
"""
|
|
return ResNet(BottleNeck, [3, 8, 36, 3])
|
|
|
|
|
|
if __name__ == '__main__':
|
|
model = ResNet(BasicBlock, [2,2,2,2])
|
|
print(model(torch.randn(2,3,32,32), torch.LongTensor([4,19])).shape)
|
|
|