-
Notifications
You must be signed in to change notification settings - Fork 14
/
resnetv2-org.py
185 lines (151 loc) · 6.29 KB
/
resnetv2-org.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
'''ResNet in PyTorch.
For Pre-activation ResNet, see 'preact_resnet.py'.
Reference:
[1] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun
Deep Residual Learning for Image Recognition. arXiv:1512.03385
'''
import torch
import torch.nn as nn
import torch.nn.functional as F
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, in_planes, planes, stride=1):
super(BasicBlock, self).__init__()
self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
self.bn1 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=1, padding=1, bias=False)
self.bn2 = nn.BatchNorm2d(planes)
self.shortcut = nn.Sequential()
if stride != 1 or in_planes != self.expansion * planes:
self.shortcut = nn.Sequential(
nn.Conv2d(in_planes, self.expansion * planes, kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(self.expansion * planes)
)
def forward(self, x):
out = F.relu(self.bn1(self.conv1(x)))
out = self.bn2(self.conv2(out))
out += self.shortcut(x)
out = F.relu(out)
return out
class Bottleneck(nn.Module):
expansion = 4
def __init__(self, in_planes, planes, stride=1):
super(Bottleneck, self).__init__()
self.conv1 = nn.Conv2d(in_planes, 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, self.expansion * planes, kernel_size=1, bias=False)
self.bn3 = nn.BatchNorm2d(self.expansion * planes)
self.shortcut = nn.Sequential()
if stride != 1 or in_planes != self.expansion * planes:
self.shortcut = nn.Sequential(
nn.Conv2d(in_planes, self.expansion * planes, kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(self.expansion * planes)
)
def forward(self, x):
out = F.relu(self.bn1(self.conv1(x)))
out = F.relu(self.bn2(self.conv2(out)))
out = self.bn3(self.conv3(out))
out += self.shortcut(x)
out = F.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, block, num_blocks, num_classes=10, zero_init_residual=False):
super(ResNet, self).__init__()
self.in_planes = 64
self.conv1 = nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1, bias=False)
self.bn1 = nn.BatchNorm2d(64)
self.layer1 = self._make_layer(block, 64, num_blocks[0], stride=1)
self.layer2 = self._make_layer(block, 128, num_blocks[1], stride=2)
self.layer3 = self._make_layer(block, 256, num_blocks[2], stride=2)
self.layer4 = self._make_layer(block, 512, num_blocks[3], stride=2)
self.avgpool = nn.AdaptiveAvgPool2d((1, 1))
self.linear = nn.Linear(512 * block.expansion, num_classes)
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
nn.init.constant_(m.weight, 1)
nn.init.constant_(m.bias, 0)
# Zero-initialize the last BN in each residual branch,
# so that the residual branch starts with zeros, and each residual block behaves like an identity.
# This improves the model by 0.2~0.3% according to https://arxiv.org/abs/1706.02677
if zero_init_residual:
for m in self.modules():
if isinstance(m, Bottleneck):
nn.init.constant_(m.bn3.weight, 0)
elif isinstance(m, BasicBlock):
nn.init.constant_(m.bn2.weight, 0)
def get_feat_modules(self):
feat_m = nn.ModuleList([])
feat_m.append(self.conv1)
feat_m.append(self.bn1)
feat_m.append(self.layer1)
feat_m.append(self.layer2)
feat_m.append(self.layer3)
feat_m.append(self.layer4)
return feat_m
def get_bn_before_relu(self):
if isinstance(self.layer1[0], Bottleneck):
bn1 = self.layer1[-1].bn3
bn2 = self.layer2[-1].bn3
bn3 = self.layer3[-1].bn3
bn4 = self.layer4[-1].bn3
elif isinstance(self.layer1[0], BasicBlock):
bn1 = self.layer1[-1].bn2
bn2 = self.layer2[-1].bn2
bn3 = self.layer3[-1].bn2
bn4 = self.layer4[-1].bn2
else:
raise NotImplementedError('ResNet unknown block error !!!')
return [bn1, bn2, bn3, bn4]
def _make_layer(self, block, planes, num_blocks, stride):
strides = [stride] + [1] * (num_blocks - 1)
layers = []
for i in range(num_blocks):
stride = strides[i]
layers.append(block(self.in_planes, planes, stride))
self.in_planes = planes * block.expansion
return nn.Sequential(*layers)
def forward(self, x, is_feat=False):
out = F.relu(self.bn1(self.conv1(x)))
f0 = out
out = self.layer1(out)
f1 = out
out = self.layer2(out)
f2 = out
out = self.layer3(out)
f3 = out
out = self.layer4(out)
f4 = out
out = self.avgpool(out)
out = out.view(out.size(0), -1)
f5 = out
out = self.linear(out)
if is_feat:
return [f0, f1, f2, f3, f4, f5], out
else:
return out
def ResNet18(**kwargs):
return ResNet(BasicBlock, [2, 2, 2, 2], **kwargs)
def ResNet34(**kwargs):
return ResNet(BasicBlock, [3, 4, 6, 3], **kwargs)
def ResNet50(**kwargs):
return ResNet(Bottleneck, [3, 4, 6, 3], **kwargs)
def ResNet101(**kwargs):
return ResNet(Bottleneck, [3, 4, 23, 3], **kwargs)
def ResNet152(**kwargs):
return ResNet(Bottleneck, [3, 8, 36, 3], **kwargs)
if __name__ == '__main__':
net = ResNet18(num_classes=100)
x = torch.randn(2, 3, 32, 32)
feats, logit = net(x, is_feat=True)
for f in feats:
print(f.shape, f.min().item())
print(logit.shape)
for m in net.get_bn_before_relu():
if isinstance(m, nn.BatchNorm2d):
print('pass')
else:
print('warning')