-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathresnet.py
231 lines (187 loc) · 7.64 KB
/
resnet.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
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
import torch
import torch.nn as nn
import torch.nn.functional as F
from utils import initialize_weights
__all__ = ['ResNet', 'resnet9', 'resnet20', 'resnet32', 'resnet44', 'resnet56', 'resnet110', 'resnetwide28x10']
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, in_planes, planes, stride=1, downsample=None):
super(BasicBlock, self).__init__()
self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
self.bn1 = nn.BatchNorm2d(in_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 != 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.conv1(self.bn1(x)))
out = self.conv2(self.bn2(out))
out += self.shortcut(x)
out = F.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, block, num_blocks, num_classes=10):
super(ResNet, self).__init__()
self.in_planes = 16
self.conv1 = nn.Sequential(nn.Conv2d(3, 16, kernel_size=3, stride=1, padding=1, bias=False), nn.BatchNorm2d(16), nn.ReLU(inplace=True))
self.layer1 = self._make_layer(block, 16, num_blocks[0], stride=1)
self.layer2 = self._make_layer(block, 32, num_blocks[1], stride=2)
self.layer3 = self._make_layer(block, 64, num_blocks[2], stride=2)
self.norm = self.norm = nn.Sequential(nn.AvgPool2d(8), nn.Flatten())
self.fc = nn.Linear(64, num_classes)
def _make_layer(self, block, planes, num_blocks, stride):
strides = [stride] + [1]*(num_blocks-1)
layers = []
for stride in strides:
layers.append(block(self.in_planes, planes, stride))
self.in_planes = planes * block.expansion
return nn.Sequential(*layers)
def forward(self, x):
out = self.conv1(x)
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
out = self.norm(out)
out = self.fc(out)
return out
def conv_bn(in_channels, out_channels, kernel_size=3, stride=1, padding=1):
list = [nn.Conv2d(in_channels, out_channels, kernel_size=kernel_size, stride=stride, padding=padding),
nn.BatchNorm2d(out_channels),
nn.CELU(alpha=0.075)]
return nn.Sequential(*list)
class Residual(nn.Module):
def __init__(self, module):
super(Residual, self).__init__()
self.module = module
def forward(self, x):
return x + self.module(x)
class ResNet9(nn.Module):
def __init__(self, num_classes: int = 10):
super(ResNet9, self).__init__()
self.conv1 = conv_bn(3, 64)
self.conv2 = conv_bn(64, 128, 5, 2, 2)
self.res1 = Residual(
nn.Sequential(
conv_bn(128, 128),
conv_bn(128, 128),
))
self.conv3 = nn.Sequential(conv_bn(128, 256),nn.MaxPool2d(2))
self.res2 = Residual(
nn.Sequential(
conv_bn(256, 256),
conv_bn(256, 256),
))
self.conv4 = nn.Sequential(conv_bn(256, 128, 3, 1, 0),
nn.AdaptiveMaxPool2d((1, 1)),
nn.Flatten())
self.fc = nn.Linear(128, num_classes, bias=False)
def forward(self, x):
out = self.conv1(x)
out = self.conv2(out)
out = self.res1(out)
out = self.conv3(out)
out = self.res2(out)
out = self.conv4(out)
out = self.fc(out)
return out
class WideBasic(nn.Module):
def __init__(self, in_planes, planes, stride=1):
super(WideBasic, self).__init__()
self.bn1 = nn.BatchNorm2d(in_planes)
self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=3, padding=1, bias=True)
self.bn2 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride, padding=1, bias=True)
self.shortcut = nn.Sequential()
if stride != 1 or in_planes != planes:
self.shortcut = nn.Sequential(
nn.Conv2d(in_planes, planes, kernel_size=1, stride=stride, bias=True),
nn.BatchNorm2d(planes)
)
def forward(self, x):
out = self.conv1(F.relu(self.bn1(x)))
out = self.conv2(F.relu(self.bn2(out)))
out += self.shortcut(x)
return out
class WideResNet(nn.Module):
def __init__(self, num_classes=10, depth=28, widen_factor=10):
super(WideResNet, self).__init__()
self.in_planes = 16
assert ((depth - 4) % 6 == 0), 'Wide-resnet depth should be 6n+4'
n = (depth - 4) / 6
k = widen_factor
nstages = [16, 16 * k, 32 * k, 64 * k]
self.conv1 = nn.Conv2d(3, nstages[0], kernel_size=3, stride=1, padding=1, bias=True)
self.layer1 = self._wide_layer(WideBasic, nstages[1], n, stride=1)
self.layer2 = self._wide_layer(WideBasic, nstages[2], n, stride=2)
self.layer3 = self._wide_layer(WideBasic, nstages[3], n, stride=2)
self.norm = nn.Sequential(nn.AvgPool2d(8), nn.Flatten())
self.fc = nn.Linear(nstages[3], num_classes)
def _wide_layer(self, block, planes, num_blocks, stride):
strides = [stride] + [1] * int(num_blocks - 1)
layers = []
for stride in strides:
layers.append(block(self.in_planes, planes, stride))
self.in_planes = planes
return nn.Sequential(*layers)
def forward(self, x):
out = self.conv1(x)
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
out = self.norm(out)
out = self.fc(out)
return out
def resnet20(num_classes=10):
net = ResNet(BasicBlock, [3, 3, 3], num_classes=num_classes)
net.embed_size = 64
net.apply(initialize_weights)
return net
def resnet32(num_classes=10):
net = ResNet(BasicBlock, [5, 5, 5], num_classes=num_classes)
net.embed_size = 64
net.apply(initialize_weights)
return net
def resnet44(num_classes=10):
net = ResNet(BasicBlock, [7, 7, 7], num_classes=num_classes)
net.embed_size = 64
net.apply(initialize_weights)
return net
def resnet56(num_classes=10):
net = ResNet(BasicBlock, [9, 9, 9], num_classes=num_classes)
net.embed_size = 64
net.apply(initialize_weights)
return net
def resnet110(num_classes=10):
net = ResNet(BasicBlock, [18, 18, 18], num_classes=num_classes)
net.embed_size = 64
net.apply(initialize_weights)
return net
def resnet9(num_classes=10):
net = ResNet9(num_classes=num_classes)
net.embed_size = 128
net.apply(initialize_weights)
return net
def resnetwide28x10(num_classes=10):
net = WideResNet(depth=28, widen_factor=10, num_classes=num_classes)
net.embed_size = 640
net.apply(initialize_weights)
return net
def test(net):
import numpy as np
total_params = 0
inp = torch.randn(size=(1, 3, 32, 32))
out = net(inp)
print(f"output: {out.size()}")
for x in filter(lambda p: p.requires_grad, net.parameters()):
total_params += np.prod(x.data.numpy().shape)
print("Total number of params", total_params)
print("Total layers", len(list(filter(lambda p: p.requires_grad and len(p.data.size())>1, net.parameters()))))
if __name__ == "__main__":
for net_name in __all__:
if net_name.startswith('resnet'):
print(net_name)
test(globals()[net_name]())