VGG16/model.py at master · YM-Byun/VGG16 · GitHub

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
import torch
import torch.nn as nn

class VGG16(nn.Module):
    def __init__(self):
        super(VGG16, self).__init__()

        self.layer1 = nn.Sequential(
            *self.make_conv_layer(in_channels=3, out_channels=64),
            *self.make_conv_layer(in_channels=64, out_channels=64),
            nn.MaxPool2d(kernel_size=2, stride=2))

        self.layer2 = nn.Sequential(
            *self.make_conv_layer(in_channels=64, out_channels=128),
            *self.make_conv_layer(in_channels=128, out_channels=128),
            nn.MaxPool2d(kernel_size=2, stride=2))

        self.layer3 = nn.Sequential(
            *self.make_conv_layer(in_channels=128, out_channels=256),
            *self.make_conv_layer(in_channels=256, out_channels=256),
            *self.make_conv_layer(in_channels=256, out_channels=256),
            nn.MaxPool2d(kernel_size=2, stride=2))

        self.layer4 = nn.Sequential(
            *self.make_conv_layer(in_channels=256, out_channels=512),
            *self.make_conv_layer(in_channels=512, out_channels=512),
            *self.make_conv_layer(in_channels=512, out_channels=512),
            nn.MaxPool2d(kernel_size=2, stride=2))

        self.layer5 = nn.Sequential(
            *self.make_conv_layer(in_channels=512, out_channels=512),
            *self.make_conv_layer(in_channels=512, out_channels=512),
            *self.make_conv_layer(in_channels=512, out_channels=512),
            nn.MaxPool2d(kernel_size=2, stride=2))

        self.conv_layers = nn.Sequential(
            self.layer1,
            self.layer2,
            self.layer3,
            self.layer4,
            self.layer5)

        self.fc = nn.Sequential(
            nn.Dropout(p=0.5),
            nn.Linear(512, 512),
            nn.LeakyReLU(inplace=True),
            nn.Dropout(p=0.5),
            nn.Linear(512, 512),
            nn.LeakyReLU(inplace=True),
            nn.Dropout(p=0.5),
            nn.Linear(512, 10))

        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='leaky_relu')

            elif isinstance(m, nn.BatchNorm2d):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)

            elif isinstance(m, nn.Linear):
                nn.init.normal_(m.weight, 0, 0.01)
                nn.init.constant_(m.bias, 0)

    def make_conv_layer(self, in_channels, out_channels, kernel_size=3, stride=1, padding=1):
        layer = []

        layer.append(
            nn.Conv2d(in_channels=in_channels,
                out_channels=out_channels,
                kernel_size=kernel_size,
                stride=stride,
                padding = padding))

        layer.append(nn.BatchNorm2d(out_channels))

        layer.append(nn.LeakyReLU(inplace=True))

        return layer

    def forward(self, x):
        x = self.conv_layers(x)
        x = x.view(x.size(0), -1)
        x = self.fc(x)

        return x


if __name__ == '__main__':
    dummy_data = torch.rand(10, 3, 32, 32)

    vgg16 = VGG16()

    print ("VGG 16 network")
    print (vgg16)

    print ("\n--------------------------------------\n")

    x = vgg16(dummy_data)

    print (f"Result: {x.shape}")