/
models.py
84 lines (71 loc) · 3.12 KB
/
models.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
## custom models for codes
## resnet codes implementation following https://github.com/kuangliu/pytorch-cifar/blob/master/models/preact_resnet.py
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
from torchvision import transforms
import numpy as np
class PreActBlock(nn.Module):
'''Pre-activation version of the BasicBlock.'''
expansion = 1
def __init__(self, in_planes, planes, stride=1, bn=True, **kwargs):
super(PreActBlock, self).__init__()
if bn:
self.bn1 = nn.BatchNorm2d(in_planes)
else:
self.bn1 = nn.Identity()
self.conv1 = nn.Conv2d(
in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
if bn:
self.bn2 = nn.BatchNorm2d(planes)
else:
self.bn2 = nn.Identity()
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3,
stride=1, padding=1, bias=False)
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)
)
def forward(self, x):
out = F.relu(self.bn1(x))
shortcut = self.shortcut(out) if hasattr(self, 'shortcut') else x
out = self.conv1(out)
out = self.conv2(F.relu(self.bn2(out)))
out += shortcut
return out
class PreActResNet(nn.Module):
def __init__(self, block, num_blocks, num_classes=10, init_channels=64, bn=True):
super(PreActResNet, self).__init__()
self.in_planes = init_channels
c = init_channels
self.conv1 = nn.Conv2d(3, c, kernel_size=3,
stride=1, padding=1, bias=False)
self.layer1 = self._make_layer(block, c, num_blocks[0], stride=1, bn=bn)
self.layer2 = self._make_layer(block, 2*c, num_blocks[1], stride=2, bn=bn)
self.layer3 = self._make_layer(block, 4*c, num_blocks[2], stride=2, bn=bn)
self.layer4 = self._make_layer(block, 8*c, num_blocks[3], stride=2, bn=bn)
self.linear = nn.Linear(8*c*block.expansion, num_classes)
def _make_layer(self, block, planes, num_blocks, stride, bn=True):
# eg: [2, 1, 1, ..., 1]. Only the first one downsamples.
strides = [stride] + [1]*(num_blocks-1)
layers = []
for stride in strides:
layers.append(block(self.in_planes, planes, stride, bn=bn))
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.layer4(out)
out = F.avg_pool2d(out, 4)
out = out.view(out.size(0), -1)
out = self.linear(out)
return out
def make_resnet18k(k=64, num_classes=10, bn=True) -> PreActResNet:
''' Returns a ResNet18 with width parameter k. (k=64 is standard ResNet18)'''
model = PreActResNet(PreActBlock, [2, 2, 2, 2], num_classes=num_classes, init_channels=k, bn=bn)
return model