-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathserver.py
222 lines (189 loc) · 8.34 KB
/
server.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
import torch
from torch import nn, optim
import torch.nn.functional as F
import time
import copy
import numpy as np
import torch
from torch import nn, optim
import torch.nn.functional as F
import time
import copy
import numpy as np
from utils import init_dict, save_dict, curve_save, time_mark, print_cz, update_lr
def avg_freq_all(
weights,
L=0.1,
is_conv=True
):
client_num = len(weights)
if is_conv:
N, C, D1, D2 = weights[0].size()
else:
N = 1
C = 1
D1, D2 = weights[0].size()
#print(N, C, D1, D2)
temp_low_amp = np.zeros((C*D1, D2*N), dtype=float)
temp_low_pha = np.zeros((C*D1, D2*N), dtype=float)
for i in range(client_num):
# N, C, D1, D2 = weights[i].size()
#weights[i] = weights[i].cpu().numpy()
if is_conv:
weights[i] = weights[i].permute(1, 2, 3, 0).reshape((C*D1, D2*N))
weights[i] = weights[i].cpu().numpy()
client_fft = np.fft.fft2(weights[i], axes=(-2, -1))
amp_fft, pha_fft = np.abs(client_fft), np.angle(client_fft)
low_part_amp = np.fft.fftshift(amp_fft, axes=(-2, -1))
temp_low_amp += low_part_amp
low_part_pha = np.fft.fftshift(pha_fft, axes=(-2, -1))
temp_low_pha += low_part_pha
temp_low_amp = temp_low_amp / 4
temp_low_pha = temp_low_pha / 4
for i in range(client_num):
client_fft = np.fft.fft2(weights[i], axes=(-2, -1))
amp_fft, pha_fft = np.abs(client_fft), np.angle(client_fft)
low_part_amp = np.fft.fftshift(amp_fft, axes=(-2, -1))
low_part_pha = np.fft.fftshift(pha_fft, axes=(-2, -1))
h, w = low_part_amp.shape
b_h = (np.floor(h *L / 2)).astype(int)
b_w = (np.floor(w *L / 2)).astype(int)
c_h = np.floor(h/2.0).astype(int)
c_w = np.floor(w/2.0).astype(int)
h1 = c_h-b_h
h2 = c_h+b_h
w1 = c_w-b_w
w2 = c_w+b_w
low_part_amp[h1:h2,w1:w2] = temp_low_amp[h1:h2,w1:w2]
low_part_amp = np.fft.ifftshift(low_part_amp, axes=(-2, -1))
low_part_pha[h1:h2,w1:w2] = temp_low_pha[h1:h2,w1:w2]
low_part_pha = np.fft.ifftshift(low_part_pha, axes=(-2, -1))
fft_back_ = low_part_amp * np.exp(1j * low_part_pha)
# get the mutated image
fft_back_ = np.fft.ifft2(fft_back_, axes=(-2, -1))
weights[i] = torch.FloatTensor(np.real(fft_back_))
if is_conv:
weights[i] = weights[i].reshape(C, D1, D2, N).permute(3, 0, 1, 2)
#print(weights[i].shape)
return weights
def avg_freq_all_finalFC(
weights,
L=0.1,
):
client_num = len(weights)
cls_num, D = weights[0].size()
print('cls_num: {:d}, D: {:d}'.format(cls_num, D))
# N = 1
# C = 1
# D1, D2 = weights[0].size() # 3*D
#print(N, C, D1, D2)
temp_low_amp = [np.zeros(D, dtype=float) for cls_idx in range(cls_num)]
temp_low_pha = [np.zeros(D, dtype=float) for cls_idx in range(cls_num)]
for i in range(client_num):
weights[i] = weights[i].cpu().numpy()
for cls_idx in range(cls_num):
client_fft = np.fft.fft(weights[i][cls_idx, :], axis=-1)
amp_fft, pha_fft = np.abs(client_fft), np.angle(client_fft)
low_part_amp = np.fft.fftshift(amp_fft, axes=(-1,))
temp_low_amp[cls_idx] += low_part_amp
low_part_pha = np.fft.fftshift(pha_fft, axes=(-1,))
temp_low_pha[cls_idx] += low_part_pha
for cls_idx in range(cls_num):
temp_low_amp[cls_idx] = temp_low_amp[cls_idx] / 4
temp_low_pha[cls_idx] = temp_low_pha[cls_idx] / 4
for i in range(client_num):
for cls_idx in range(cls_num):
client_fft = np.fft.fft(weights[i][cls_idx, :], axis=-1)
amp_fft, pha_fft = np.abs(client_fft), np.angle(client_fft)
low_part_amp = np.fft.fftshift(amp_fft, axes=(-1,))
low_part_pha = np.fft.fftshift(pha_fft, axes=(-1,))
h = low_part_amp.shape[0]
b_h = (np.floor(h *L / 2)).astype(int)
c_h = np.floor(h/2.0).astype(int)
h1 = c_h-b_h
h2 = c_h+b_h
low_part_amp[h1:h2] = temp_low_amp[cls_idx][h1:h2]
low_part_amp = np.fft.ifftshift(low_part_amp, axes=(-1,))
low_part_pha[h1:h2] = temp_low_pha[cls_idx][h1:h2]
low_part_pha = np.fft.ifftshift(low_part_pha, axes=(-1,))
fft_back_ = low_part_amp * np.exp(1j * low_part_pha)
# get the mutated image
fft_back_ = np.fft.ifft(fft_back_, axis=-1)
weights[i][cls_idx, :] = np.real(fft_back_)
#print(weights[i].shape)
#
weights[i] = torch.FloatTensor(weights[i])
return weights
def PFA(
weights,
L,
is_conv
):
return avg_freq_all(weights=weights, L=L, is_conv=is_conv)
def PFA_finalFC(
weights,
L
):
return avg_freq_all_finalFC(weights=weights, L=L)
################# Key Function ########################
def communication(
args,
server_model,
models,
original_models,
client_weights,
a_iter
):
pfa_rate = args.l_rate + (a_iter / args.iters) * (0.95 - args.l_rate)
client_num = len(client_weights) #
with torch.no_grad():
# aggregate params
for key in server_model.state_dict().keys():
if 'bn' not in key: #not bn
if 'conv' in key and 'weight' in key:
temp_weights = PFA(
[
models[0].state_dict()[key].data,
models[1].state_dict()[key].data,
models[2].state_dict()[key].data,
models[3].state_dict()[key].data
],
L=pfa_rate,
is_conv=True
)
for client_idx in range(client_num): # copy from server to each client
models[client_idx].state_dict()[key].data.copy_(temp_weights[client_idx])
elif 'linear' in key and 'weight' in key and 'fc2' not in key:
temp_weights = PFA(
[
models[0].state_dict()[key].data,
models[1].state_dict()[key].data,
models[2].state_dict()[key].data,
models[3].state_dict()[key].data
],
L=pfa_rate,
is_conv=False
)
for client_idx in range(client_num): #
models[client_idx].state_dict()[key].data.copy_(temp_weights[client_idx])
elif 'linear' in key and 'weight' in key and 'fc2' in key:
temp_weights = PFA_finalFC(
[
models[0].state_dict()[key].data,
models[1].state_dict()[key].data,
models[2].state_dict()[key].data,
models[3].state_dict()[key].data
],
L=pfa_rate,
)
for client_idx in range(client_num): #
models[client_idx].state_dict()[key].data.copy_(temp_weights[client_idx])
else:
print(key, '\t not bn, conv, fc layer, with param!')
temp = torch.zeros_like(server_model.state_dict()[key], dtype=torch.float32)
for client_idx in range(client_num):
temp += client_weights[client_idx] * models[client_idx].state_dict()[key]
server_model.state_dict()[key].data.copy_(temp) # non-bn layer，update the server model
for client_idx in range(client_num): # non-bn layer, from server to each client
models[client_idx].state_dict()[key].data.copy_(server_model.state_dict()[key])
return server_model, models