Repository navigation
Expand file tree
/
Copy pathmodel.py
More file actions
96 lines (74 loc) · 3.47 KB
/
Copy pathmodel.py
File metadata and controls
96 lines (74 loc) · 3.47 KB
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
import torch.nn as nn
import torch.nn.init as init
from myconv import myconv2d
from mask import *
randommask = 1
class Lambda(nn.Module):
def __init__(self, func):
super().__init__()
self.func = func
def forward(self, x): return self.func(x)
def mnist_resize(x): return x.view(-1, 1, 28, 28)
def cifar_resize(x): return x.view(-1, 3, 32, 32)
def imagenet_resize(x): return x.view(-1, 3, 128, 128)
def noop(x): return x
class Flatten(nn.Module):
def forward(self, x): return x.view(x.size(0), -1)
act_func = nn.ReLU()
first = True
def conv(ni, no, ks, s=1, bias=False,mask_layer=False):
return nn.Conv2d(ni, no, kernel_size=ks, stride=s, padding=ks//2, bias=bias)
def conv_layer(ni, no, ks, s, zero_bn=False, act=True,mask_layer=True):
bn = nn.BatchNorm2d(no)
nn.init.constant_(bn.weight, 0. if zero_bn else 1.)
layers = [conv(ni, no, ks=ks, s=s,mask_layer=mask_layer), bn]
if act: layers += [act_func]
return nn.Sequential(*layers)
def init_cnn(m):
if isinstance(m, (nn.Conv2d, nn.Linear)): init.kaiming_normal_(m.weight, a=1.0)
if getattr(m, 'bias', None) is not None: init.constant_(m.bias, 0.)
for l in m.children(): init_cnn(l)
class ResBlock(nn.Module):
def __init__(self, ni, no, expansion, s=1):
super().__init__()
ni *= expansion
layers = [conv_layer(ni, no, 3, s),
conv_layer(no, no*expansion, 3, 1, zero_bn=True, act=False)
] if expansion == 1 else [
conv_layer(ni, no, 1, 1),
conv_layer(no, no, 3, s),
conv_layer(no, no*expansion, 1, 1, zero_bn=True, act=False)
]
self.convs = nn.Sequential(*layers)
self.idconv = noop if ni == no*expansion else conv_layer(ni, no*expansion, 1, 1, act=False)
self.pool = noop if s == 1 else nn.AvgPool2d(2,ceil_mode=True)
def forward(self, x):
return act_func(self.convs(x) + self.idconv(self.pool(x)))
class XResNet(nn.Sequential):
@classmethod
def create(cls, expansion, layers, c_in=3, c_out=10,resize=cifar_resize,compression_factor=1):
nbs = [c_in, 32,64,64]
stem = [conv_layer(nbs[i], nbs[i+1], 3, 2 if i==0 else 1,False)
for i in range(3)]
maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
nbs = [64//expansion, 64, 128//compression_factor, 256//compression_factor, 512//compression_factor]
res_layers = [cls._make_layer(nbs[i], nbs[i+1], expansion, 1 if i==0 else 2, l)
for i,l in enumerate(layers) ]
layers = [Lambda(resize), *stem, maxpool, *res_layers]
layers.extend([nn.AdaptiveAvgPool2d(1), Flatten(), nn.Linear(nbs[-1]*expansion, c_out) ])
resnet = cls(*layers)
init_cnn(resnet)
return resnet
@staticmethod
def _make_layer(ni, no, expansion, s, number_block):
return nn.Sequential(*[ResBlock(ni if i==0 else no, no, expansion, s=s if i==0 else 1)
for i in range(number_block)] )
def xresnet18 (mask=1,**kwargs):
global randommask
randommask = mask
return XResNet.create(1, [2, 2, 2, 2], **kwargs)
def xresnet34 (**kwargs): return XResNet.create(1, [3, 4, 6, 3], **kwargs)
def xresnet50 (**kwargs): return XResNet.create(4, [3, 4, 6, 3], **kwargs)
def xresnet100 (**kwargs): return XResNet.create(4, [6, 8, 12, 6], **kwargs)
def xresnet101(**kwargs): return XResNet.create(4, [3, 4, 23, 3], **kwargs)
def xresnet152(**kwargs): return XResNet.create(4, [3, 8, 36, 3], **kwargs)