-
Notifications
You must be signed in to change notification settings - Fork 2
/
resnet.py
266 lines (229 loc) · 10 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
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from torch.utils.data import random_split
import torch.utils.data as Data
import torchvision
import torchvision.transforms as transforms
import torch.optim as optim
import scipy.io as sio
# 用于ResNet18和34的残差块,用的是2个3x3的卷积
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()
# 经过处理后的x要与x的维度相同(尺寸和深度)
# 如果不相同,需要添加卷积+BN来变换为同一维度
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
# 用于ResNet50,101和152的残差块,用的是1x1+3x3+1x1的卷积
class Bottleneck(nn.Module):
# 前面1x1和3x3卷积的filter个数相等,最后1x1卷积是其expansion倍
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, linear_size=32):
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.linear = nn.Linear(int((linear_size ** 2) / 2) * block.expansion, 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 = F.relu(self.bn1(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 ResNet18(linear_size=32):
return ResNet(BasicBlock, [2, 2, 2, 2], linear_size=linear_size)
def ResNet34():
return ResNet(BasicBlock, [3, 4, 6, 3])
def ResNet50():
return ResNet(Bottleneck, [3, 4, 6, 3])
def ResNet101():
return ResNet(Bottleneck, [3, 4, 23, 3])
def ResNet152():
return ResNet(Bottleneck, [3, 8, 36, 3])
"""
https://blog.csdn.net/qq_36370187/article/details/103103382
"""
def PretreatmentData():
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
'''
transforms.Compose 这个类的主要作用是串联多个图片变换的操作
transforms.RandomCrop 裁剪大小
RandomHorizontalFlip 水平翻转
Normalize 标准化
'''
# dataset=loadCIFAR()
# dataset = loadMNIST()
dataset, label_size, data_size = loadSTL10()
# label_size=dataset.tensors[0].shape[2]
print(label_size)
train_data, eval_data = random_split(dataset, [round(0.05 * data_size),
round(0.95 * data_size)],
generator=torch.Generator().manual_seed(42)) # 把数据机随机切分训练集和验证集
print(len(train_data))
print(f"train data size is {round(0.05 * data_size)}")
print(f"test data size is {round(0.95 * data_size)}")
train_loader = Data.DataLoader(dataset=train_data, batch_size=50, shuffle=True, num_workers=2, drop_last=False)
test_loader = Data.DataLoader(dataset=eval_data, batch_size=500, shuffle=False, num_workers=2)
net = ResNet18(linear_size=label_size).to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.01, momentum=0.9, weight_decay=5e-4)
net.train()
for epoch in range(0, 100):
print(f'\n EPOCH: {epoch + 1}')
sum_loss = 0.0
correct = 0.0
total = 0.0
for i, data in enumerate(train_loader, 0):
length = len(train_loader)
inputs, labeles = data
inputs, labeles = inputs.to(device), labeles.to(device)
optimizer.zero_grad() # 梯度归零
outputs = net(inputs)
loss = criterion(outputs, labeles) # 交叉熵损失函数
loss.backward() # 反向传播计算梯度值
optimizer.step() # 梯度下降,更新参数值
sum_loss += loss.item()
_, predicted = torch.max(outputs.data, 1)
total += labeles.size(0)
correct += predicted.eq(labeles.data).cpu().sum()
print(
f'[epoch:{epoch + 1},iter:{i + 1 + epoch * length}] Loss:{sum_loss / (i + 1):.3f} '
f'|Acc:{100. * correct / total:.3f}')
if sum_loss <= .01:
print(f"train stop when loss == {sum_loss} ,epoch:{epoch + 1},iter:{i + 1 + epoch * length}")
break
if sum_loss <= .01:
break
print('Waiting Test...')
with torch.no_grad():
correct = 0
total = 0
for i, data in enumerate(test_loader, 0):
net.eval() # 固定参数
images, labeles = data
images, labeles = images.to(device), labeles.to(device)
outputs = net(images)
_, predicted = torch.max(outputs.data, 1)
total += labeles.size(0)
tmp_corr = (predicted == labeles).sum()
print(f'Test {i} ac is ------->{100 * tmp_corr / labeles.size(0):.3f}')
correct += tmp_corr
print(f'Test ac is:{100 * correct / total:.3f}')
def loadSTL10():
from sklearn.preprocessing import LabelEncoder
a = sio.loadmat("./data_set/test.mat")
data = a['X']
target1 = a['y']
data1 = np.reshape(data, (8000, 3, 96, 96))
data1 = np.transpose(data1, [0, 3, 2, 1]) # 更改维度,
b = sio.loadmat("./data_set/train.mat")
data = b['X']
target2 = b['y']
data2 = np.reshape(data, (-1, 3, 96, 96))
data2 = np.transpose(data2, [0, 3, 2, 1])
data = np.concatenate((data1, data2)) # 合并数据
target = LabelEncoder().fit_transform(np.squeeze(np.concatenate((target1, target2))))
transform = transforms.Compose([
transforms.ToTensor(),
transforms.Resize(64),
transforms.RandomHorizontalFlip(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010))
])
res = []
for d in data:
res.append(transform(d))
data = torch.stack(res)
print(target.shape)
return Data.TensorDataset(data.float(), torch.tensor(target).long()), 64, target.shape[0]
def loadCIFAR():
transform = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010))
])
trainset = torchvision.datasets.CIFAR10(root="./data", train=True, download=True, transform=transform)
testset = torchvision.datasets.CIFAR10(root='./data', train=False, download=True, transform=transform)
train = np.transpose(np.concatenate((trainset.data, testset.data)), [0, 3, 1, 2])
target = trainset.targets + testset.targets
dataset = Data.TensorDataset(torch.tensor(train).float(), torch.tensor(target).long())
return dataset, 32, len(target)
def loadMNIST():
transform = transforms.Compose([
#transforms.ToPILImage,
transforms.Grayscale(num_output_channels=3), # 转为RGB图像
transforms.Resize([32, 32]),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010))
])
trainset = torchvision.datasets.MNIST(root="./data", train=True, download=True, transform=transform)
# print(trainset.data.shape)
# res=[]
# for d in trainset.data:
# res.append(transform(d))
# data=torch.stack(res)
# print(data.shape)
# return Data.TensorDataset(trainset.data,trainset.targets)
return trainset, 32, len(trainset.targets)
if __name__ == '__main__':
PretreatmentData()