-
Notifications
You must be signed in to change notification settings - Fork 3
/
Communication_GAN.py
161 lines (148 loc) · 7.25 KB
/
Communication_GAN.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
import os
from data_utils import generate_partial_femnist,pre_handle_femnist_mat,get_mnist_dataset,get_device_num,get_device_id
from models import CNN_3layer_fc_model,CNN_2layer_fc_model,CNN_3layer_fc_model_no_softmax,CNN_2layer_fc_model_no_softmax,DomainIdentifierNew
from collaborate_train import train_models_bal_femnist_collaborate,feature_domain_alignment,train_models_collaborate_gan
from model_utils import get_model_list,get_femnist_model_list,test_models_femnist
from matplotlib.pyplot import MultipleLocator #从pyplot导入MultipleLocator类，这个类用于设置刻度间隔
from models import DomainIdentifier
import matplotlib.pyplot as plt
from option import args_parser
import mindspore
from mindspore import save_checkpoint
from mindspore.communication.management import init
from mindspore import context
def transpose( matrix):
"""
Matrix transpose
"""
new_matrix = []
for i in range(len(matrix[0])):
matrix1 = []
for j in range(len(matrix)):
matrix1.append(matrix[j][i])
new_matrix.append(matrix1)
return new_matrix
if __name__ == '__main__':
"""
Init context
"""
args = args_parser()
# device ='cuda' if args.gpu else 'cpu'
mindspore.context.set_context(mode=mindspore.context.GRAPH_MODE, device_target=args.device_target)
device = 'Ascend'
context.set_context(mode=context.GRAPH_MODE,
device_target=args.device_target)
device_num = get_device_num()
if args.device_target == "Ascend":
device_id = get_device_id()
context.set_context(device_id=device_id)
if device_num > 1:
context.reset_auto_parallel_context()
context.set_auto_parallel_context(device_num=device_num, parallel_mode='data_parallel',
gradients_mean=True)
init()
"""
Init models
"""
datasetindex = 0
models = {"2_layer_CNN": CNN_2layer_fc_model,"3_layer_CNN": CNN_3layer_fc_model}
models_no_softmax = {"2_layer_CNN": CNN_2layer_fc_model_no_softmax,"3_layer_CNN": CNN_3layer_fc_model_no_softmax}
models_ini_list = [{"model_type": "2_layer_CNN", "params": {"n1": 128, "n2": 256, "dropout_rate": 0.2}},
{"model_type": "2_layer_CNN", "params": {"n1": 128, "n2": 384, "dropout_rate": 0.2}},
{"model_type": "2_layer_CNN", "params": {"n1": 128, 'n2': 512, "dropout_rate": 0.2}},
{"model_type": "2_layer_CNN", "params": {"n1": 256, "n2": 256, "dropout_rate": 0.3}},
{"model_type": "2_layer_CNN", "params": {"n1": 256, "n2": 512, "dropout_rate": 0.4}}]
accuracy = []
models_list = []
eachround_accuracy = []
root = "./Model"
name = ["Femnist_model_0.ckpt", "Femnist_model_1.ckpt", "Femnist_model_2.ckpt",
"Femnist_model_3.ckpt", "Femnist_model_4.ckpt"]
models_list = get_model_list(root=root, name=name, models_ini_list=models_ini_list, models=models)
"""
Generate femnist dataset
"""
X_train, y_train, writer_ids_train, X_test, y_test, writer_ids_train, writer_ids_test = pre_handle_femnist_mat()
y_train += len(args.public_classes)
y_test += len(args.public_classes)
femnist_X_test, femnist_y_test = generate_partial_femnist(X=X_test, y=y_test, class_in_use=args.private_classes,
verbose=False)
"""
Test the models on femnist
"""
figureurl = 'Figures/mnist/collaborate_gan/'
eachround_accuracy = test_models_femnist(device=device, models_list=models_list, test_x=femnist_X_test,
test_y=femnist_y_test, savelurl=figureurl)
accuracy.append(eachround_accuracy)
"""
Define GanModel0
"""
net = DomainIdentifier()
save_checkpoint(net, 'GanModel0.ckpt')
"""
federated communication
"""
for item in range(args.Communicationepoch):
print('This is {} time communication'.format(item))
'''
Generate MNIST
'''
mnist_data_train, mnist_data_validation, mnist_data_test = get_mnist_dataset()
'''
For Latent Embedding Adaptation
'''
root = "./Model"
name = ["Femnist_model_0.ckpt", "Femnist_model_1.ckpt", "Femnist_model_2.ckpt",
"Femnist_model_3.ckpt", "Femnist_model_4.ckpt"]
models_list = get_model_list(root=root, name=name, models_ini_list=models_ini_list, models=models)
feature_domain_alignment(device=device, train=mnist_data_train,
models_list=models_list, modelurl=root,
domain_identifier_epochs=args.Communication_domain_identifier_epochs,
gan_local_epochs=args.Communication_gan_local_epochs)
'''
model agnostic federated learning
'''
root = "./Model/collaborate_gan"
name = ["LocalModel0.ckpt", "LocalModel1.ckpt", "LocalModel2.ckpt",
"LocalModel3.ckpt", "LocalModel4.ckpt"]
models_list= get_model_list(root=root,name=name,models_ini_list=models_ini_list, models=models_no_softmax)
train_models_collaborate_gan(device=device, models_list=models_list,
train=mnist_data_train, user_number=args.user_number,
collaborative_epoch=args.collaborative_epoch,output_classes=args.output_classes)
root = "./Model/final_model"
name = ["LocalModel0.ckpt", "LocalModel1.ckpt", "LocalModel2.ckpt",
"LocalModel3.ckpt", "LocalModel4.ckpt"]
models_list= get_model_list(root=root,name=name,models_ini_list=models_ini_list, models=models_no_softmax)
train_models_bal_femnist_collaborate(device=device, models_list=models_list,modelurl=root)
"""
Get the updated models
"""
models_list= get_model_list(root=root,name=name,models_ini_list=models_ini_list, models=models_no_softmax)
"""
Create test dataset
"""
X_train, y_train, writer_ids_train, X_test, y_test, writer_ids_train, writer_ids_test = pre_handle_femnist_mat()
y_train += len(args.public_classes)
y_test += len(args.public_classes)
femnist_X_test, femnist_y_test = generate_partial_femnist(X=X_test, y=y_test, class_in_use=args.private_classes, verbose=False)
figureurl = 'Figures/mnist/collaborate_gan/'
eachround_accuracy = test_models_femnist(device=device, models_list=models_list, test_x=femnist_X_test,
test_y=femnist_y_test, savelurl=figureurl)
accuracy.append(eachround_accuracy)
print(accuracy)
accuracy = transpose(accuracy)
for i, val in enumerate(accuracy):
print(val)
plt.plot(range(len(val)), val, label='model :' + str(i))
plt.legend(loc='best')
plt.title('communication_round_with_accuracy_on_femnist')
plt.xlabel('Communication round')
plt.ylabel('Accuracy on FEMNIST')
x_major_locator = MultipleLocator(1)
ax = plt.gca()
ax.xaxis.set_major_locator(x_major_locator)
plt.xlim(0, args.Communicationepoch)
figureurl = 'Figures/mnist/collaborate_gan/'
plt.savefig(figureurl + 'communication_round_with_accuracy_on_femnist.png')
plt.show()
print('End')