-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathTSModel.py
177 lines (158 loc) · 7.01 KB
/
TSModel.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
from model.DNN import DNNNetOne, DNNNetTwo
import torch.optim as optim
import torch.nn as nn
import torch.nn.functional as F
import torch
from model import Dataset as DS
import os
import pickle
seed = 9999
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
# train teacher
class Teacher:
def __init__(self, input_size, hidden_size, num_classes, device):
self.input_size = input_size
self.hidden_size = hidden_size
self.num_classes = num_classes
self.device = device
def train(self, trainloader, epoch_num):
# set model to training mode
net = DNNNetOne(self.input_size, self.hidden_size, self.num_classes)
net = net.to(self.device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(net.parameters(),
lr=0.001, betas=(0.9, 0.999), eps=1e-08,
weight_decay=0, amsgrad=False)
net.train()
for epoch in range(epoch_num): # loop over the dataset multiple times
running_loss = 0.0
for i, data in enumerate(trainloader, 0):
# get the inputs; data is a list of [inputs, labels]
inputs, labels = data
inputs = inputs.to(self.device)
labels = labels.to(self.device)
# zero the parameter gradients
optimizer.zero_grad()
# forward + backward + optimize
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
# print statistics
running_loss += loss.item()
if i % 2000 == 1999: # print every 2000 mini-batches
print('[%d, %5d] loss: %.3f' %
(epoch + 1, i + 1, running_loss / 2000))
running_loss = 0.0
return net
# use teacher to label the data
class TeacherLabel:
def __init__(self, model):
self.model = model
def label(self, x):
self.model.eval()
predict = self.model(x)
return predict
# train student
class Student:
def __init__(self, input_size1, input_size2, hidden_size, num_classes, device):
self.input_size1 = input_size1
self.input_size2 = input_size2
self.hidden_size = hidden_size
self.num_classes = num_classes
self.device = device
def train(self, trainloader, teacher1, teacher2, alpha,beta, T, epoch_num):
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
# set model to training mode
net = DNNNetTwo(self.input_size1, self.input_size2, self.hidden_size, self.num_classes)
net = net.to(self.device)
optimizer = optim.Adam(net.parameters(),
lr=0.001, betas=(0.9, 0.999), eps=1e-08,
weight_decay=0, amsgrad=False)
criterion = nn.CrossEntropyLoss()
net.train()
for epoch in range(epoch_num): # loop over the dataset multiple times
running_loss = 0.0
for i, data in enumerate(trainloader, 0):
# get the inputs; data is a list of [inputs, labels]
x1, x2, labels = data
x1 = x1.to(self.device)
x2 = x2.to(self.device)
labels = labels.to(self.device)
# teachers label
soft_label1 = TeacherLabel(teacher1).label(x1)
q1 = F.softmax(soft_label1 / T, dim=1)
soft_label2 = TeacherLabel(teacher2).label(x2)
q2 = F.softmax(soft_label2 / T, dim=1)
q1 = q1.detach()
q2 = q2.detach()
# zero the parameter gradients
optimizer.zero_grad()
# forward + backward + optimize
outputs = net(x1, x2)
p = F.log_softmax(outputs / T, dim=1)
loss1 = criterion(outputs, labels)
loss2 = nn.KLDivLoss(reduction='batchmean')(p, q1) * (T * T * alpha)
loss3 = nn.KLDivLoss(reduction='batchmean')(p, q2) * (T * T * beta)
loss = loss1 + loss2 + loss3
loss.backward()
optimizer.step()
# print statistics
running_loss += loss.item()
# print('[%d] loss: %.3f' %
# (epoch + 1, running_loss / (i+1)))
T1_W = teacher1.fc1.weight.data
T2_W = teacher2.fc1.weight.data
T1_W_sum = torch.sum(T1_W * T1_W, dim=0)
T1_W_sum = F.normalize(T1_W_sum, p=2, dim=0)
T2_W_sum = torch.sum(T2_W * T2_W, dim=0)
T2_W_sum = F.normalize(T2_W_sum, p=2, dim=0)
S1_W = net.fc11.weight
S2_W = net.fc21.weight
S1_W_sum = torch.sum(S1_W * S1_W, dim=0)
S1_W_sum = F.normalize(S1_W_sum, p=2, dim=0)
S2_W_sum = torch.sum(S2_W * S2_W, dim=0)
S2_W_sum = F.normalize(S2_W_sum, p=2, dim=0)
return net
class TeacherStudent:
def __init__(self, nclasses, ninput1, nhidden1, ninput2, nhidden2, nhidden, device, model_dir):
self.nclasses = nclasses
self.ninput1 = ninput1
self.nhidden1 = nhidden1
self.ninput2 = ninput2
self.nhidden2 = nhidden2
self.nhidden = nhidden
self.device = device
self.model_dir = model_dir
def train(self, trainloader1, trainloader2, x1, x2, y, alpha, beta,
temperature, batch_size, epoch_num):
if not os.path.isfile(self.model_dir + 'Te1.pkl'):
Mt1 = Teacher(self.ninput1, self.nhidden1, self.nclasses,
self.device) # teacher model 1
teacher1 = Mt1.train(trainloader1, epoch_num)
with open(self.model_dir + 'Te1.pkl', 'wb') as output:
pickle.dump({'Te1': teacher1}, output, pickle.HIGHEST_PROTOCOL)
else:
with open(self.model_dir + 'Te1.pkl', 'rb') as input:
Te1 = pickle.load(input)
teacher1 = Te1['Te1']
if not os.path.isfile(self.model_dir + 'Te2.pkl'):
Mt2 = Teacher(self.ninput2, self.nhidden2, self.nclasses,
self.device) # teacher model 2
teacher2 = Mt2.train(trainloader2, epoch_num)
with open(self.model_dir + 'Te2.pkl', 'wb') as output:
pickle.dump({'Te2': teacher2}, output, pickle.HIGHEST_PROTOCOL)
else:
with open(self.model_dir + 'Te2.pkl', 'rb') as input:
Te2 = pickle.load(input)
teacher2 = Te2['Te2']
trainset = DS.DatasetTwo(x1, x2, y)
trainloader = torch.utils.data.DataLoader(trainset,
batch_size=batch_size,
shuffle=True, num_workers=1)
Ms = Student(self.ninput1, self.ninput2, self.nhidden, self.nclasses,
self.device) # student model
student = Ms.train(trainloader, teacher1, teacher2, alpha, beta, temperature, epoch_num)
return student, teacher1, teacher2