federatedSpeechCommands/federated_utils_gpu.py at master · zhanglei1949/federatedSpeechCommands · GitHub

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
import torch, time, threading
import numpy as np
import math
import random
from scipy.linalg import null_space
def getLenOfGradientVector(current_grad):
    #expect a list consists of numpy arrays
    n = 0
    for arr in current_grad:
        #print(list(arr.view(-1,).shape)[0])
        n += list(arr.view(-1,).shape)[0]
    return n
def getShapeList(current_grad):
    # return the list of shapes of grad vectors, for recover
    res = []
    for arr in current_grad:
        res.append(arr.shape)
    return res
def transListOfArraysToArrays(current_grad, n):
    # expect a list of arrays, return a squeezed array, n is the total length
    res = np.zeros((n))
    ind = 0
    for arr in current_grad:
        arr = arr.view(-1,).cpu()
        res[ind:ind+arr.shape[0]] = arr
        ind+=arr.shape[0]
    return res
def listMulti(l):
    res = 1
    for ele in l:
        res *= ele
    return res
def transNumpyGrad2Cuda(grad, shape_list):
    res = []
    ind = 0
    for shape in shape_list:
        tmp = torch.from_numpy(grad[ind:ind+listMulti(shape)])
        tmp = tmp.view(shape)
        res.append(tmp.float().cuda())
        ind += listMulti(shape)
    return res

############ gpu version
def getLenOfGradientVectorCuda(current_grad):
    #expect a list consists of numpy arrays
    n = 0
    for arr in current_grad:
        #print(list(arr.view(-1,).shape)[0])
        n += list(arr.view(-1,).shape)[0]
    return n
def getShapeListCuda(current_grad):
    # return the list of shapes of grad vectors, for recover
    res = []
    for arr in current_grad:
        res.append(arr.shape)
    return res
def get_shape_and_length_gradient_cuda(current_grad):
    n = 0
    res = []
    for arr in current_grad:
        n += list(arr.view(-1,).shape)[0]
        res.append(arr.shape)
    return n, res
def transListOfArraysToArraysCuda(current_grad, n):
    # expect a list of arrays, return a squeezed cuda array, n is the total length
    res = torch.zeros(n).cuda()
    ind = 0
    for arr in current_grad:
        arr = arr.view(-1,)
        res[ind:ind+arr.shape[0]] = arr
        ind+=arr.shape[0]
    return res
def listMultiCuda(l):
    res = 1
    for ele in l:
        res *= ele
    return res
def transCudaArrayWithShapeList(grad, shape_list):
    res = []
    ind = 0
    for shape in shape_list:
        tmp = grad[ind:ind+listMulti(shape)]
        tmp = tmp.view(shape)
        res.append(tmp.float())
        ind += listMulti(shape)
    return res

class Federated:
    def __init__(self, num_clients, matrix_size, num_threads):
        # utilize a sample gradient vector
        self.num_clients = num_clients
        self.matrix_size = matrix_size
        self.num_threads = num_threads
        self.MAX = 0.001
        self.S_i = random.sample(range(0, 3 * self.matrix_size), self.matrix_size)
        self.S_i.sort()
        self.S_j = random.sample(range(0, 2 * self.matrix_size), self.matrix_size)
        self.S_j.sort()
    def init(self, gradient, shape_list):
        self.len_gradient = list(gradient.shape)[0]
        self.shape_list = shape_list
        #print("gradient length", self.len_gradient)
        #self.len_gradient, self.shape_list = get_shape_and_length_gradient_cuda(gradient)
        self.len_gradient_after_padding = math.ceil(float(self.len_gradient) / (self.matrix_size * self.num_threads)) * self.matrix_size * self.num_threads

        self.ori_gradient_sum = torch.zeros(self.len_gradient).cuda()
        #print(self.ori_gradient_sum.shape)
        self.random_gradient_sum = torch.zeros(self.len_gradient_after_padding * 3).cuda()

        self.A = self.MAX * torch.rand(self.matrix_size, self.matrix_size).float().cuda()
        self.A_inv = self.A.inverse()
        self.B = torch.zeros(self.matrix_size, 3 * self.matrix_size).cuda()
        for i in range(0, self.matrix_size):
            self.B[:, self.S_i[i] : self.S_i[i]+1] = self.A[:, i:i+1]
        self.C = (torch.rand(2 * self.matrix_size, 3 * self.matrix_size) * self.MAX).float().cuda()
        for i in range(0, self.matrix_size):
            self.C[self.S_j[i] : self.S_j[i] + 1, :] = self.B[i:i+1 , :]

        # SVD
        self.u, self.s, self.vh = torch.svd(self.C, some=False)
        self.vh_t = self.vh.t()
        self.sigma = torch.zeros(self.C.shape[0], self.C.shape[1]).cuda()
        self.sigma[: self.s.shape[0], : self.s.shape[0]] = self.s.diag()
        # null space
        self.u_sigma = torch.mm(self.u, self.sigma)
        ns = null_space(self.C.cpu()) # (3000, 1000) we use the first args.
        self.ns = torch.from_numpy(ns).cuda()
        self.trans_i = torch.zeros(self.matrix_size, 3*self.matrix_size).cuda()
        self.trans_j = torch.zeros(self.matrix_size, 2*self.matrix_size).cuda()
        for i,ind in  enumerate(self.S_i):
            self.trans_i[i][ind] = 1
        for i,ind in enumerate(self.S_j):
            self.trans_j[i][ind] = 1
        #print("initialization complete")
    def work_for_client(self, client_no, gradient):
        assert(client_no < self.num_clients)
        part_num = self.len_gradient_after_padding / self.num_threads
        time1 = time.time()
        ### TODO time issue
        flatterned_grad = gradient
        self.ori_gradient_sum += flatterned_grad
        # padding
        flatterned_grad_extended = torch.zeros(self.len_gradient_after_padding, 1).cuda()
        flatterned_grad_extended[:self.len_gradient, 0] = flatterned_grad
        kernel_space = torch.zeros(3 * self.matrix_size, 1).cuda()
        random_numbers = self.MAX * torch.rand(self.matrix_size, 1).cuda()
        ###TODO
        #kernel_space = torch.zeros(self.matrix_size, 1).cuda()
        for i in range(self.matrix_size):
            kernel_space += random_numbers[i] * self.ns[:, i:i+1]
        #print("kernel space complete")
        flatterned_grad_extended_after_random = (self.MAX * torch.rand(3 * self.len_gradient_after_padding, 1)).float().cuda()

        def randomizing_matrix(thread_id, part_num):
            for i in range(int(thread_id * part_num), int((thread_id + 1) * part_num), self.matrix_size):
                #if (thread_id == 0):
                #    print(thread_id, i/self.matrix_size, (int((thread_id + 1) * part_num)/self.matrix_size))
                flatterned_grad_extended_after_random[i * 3 : 3 * (i + self.matrix_size)] = \
                    (torch.mm(flatterned_grad_extended[i : i + self.matrix_size].view(1, self.matrix_size), \
                        self.trans_i)).view(3*self.matrix_size,1)
        threads = []
        for _i in range(self.num_threads):
            t = threading.Thread(target = randomizing_matrix, args = (_i, part_num))
            threads.append(t)
            t.start()
        for thread in threads:
            thread.join()
        # compute result
        time2 = time.time()
        #print("client ", client_no, " randomization complete ", time2 - time1)
        flatterned_grad_extended_final = (self.MAX * torch.rand(3 * self.len_gradient_after_padding, 1)).float().cuda()
        ###TODO: multi threading
        def matrixProd(thread_id, part_num):

            for i in range(int(thread_id * part_num), int((thread_id + 1) * part_num), self.matrix_size):
                flatterned_grad_extended_final[3 * i : 3*(i + self.matrix_size), :] \
                    = torch.mm(self.vh_t, flatterned_grad_extended_after_random[3 * i : 3*(i + self.matrix_size), :] + kernel_space)
            #print(thread_id, " finish")
        threads = []
        for _i in range(self.num_threads):
            t = threading.Thread(target = matrixProd, args = (_i, part_num))
            threads.append(t)
            t.start()
        for thread in threads:
            thread.join()

        #for i in range(0, self.len_gradient_after_padding, self.matrix_size):
        #    flatterned_grad_extended_final[3 * i : 3*(i + self.matrix_size), :] = torch.mm(self.vh_t, flatterned_grad_extended_after_random[3 * i : 3*(i + self.matrix_size), :] + kernel_space)

        self.random_gradient_sum += flatterned_grad_extended_final[:, 0]
        time3 = time.time()
        #self.random_gradient_sum += torch.randn(3 * self.len_gradient_after_padding).cuda() * 0.0001
        #print("client ",  client_no, " masking complete",)
        #print("time for masking", time3 - time2)
    def recoverGradient(self):
        time1 = time.time()
        res = torch.zeros(self.len_gradient_after_padding, 1).cuda()
        alpha = torch.zeros(self.matrix_size, 1).cuda()

        for i in range(0, self.len_gradient_after_padding * 3 , 3 * self.matrix_size):
            ###TODO: time issue
            tmp = torch.mm(self.u_sigma, self.random_gradient_sum[i : i + 3 * self.matrix_size].view(-1,1))
            alpha = torch.mm(self.trans_j, tmp)
            res[int(i/3) : int(i/3) + self.matrix_size, :] = torch.mm(self.A_inv, alpha)
        # set the gradient manually and update
        recovered_grad_in_cuda = transCudaArrayWithShapeList(res, self.shape_list)
        #print('[dist]\t', torch.dist(res[:self.len_gradient], self.ori_gradient_sum))
        time2 = time.time()
        #print('[dist]\t', torch.sum(torch.abs(res[:self.len_gradient, 0] - self.ori_gradient_sum)))
        #self.ori_gradient_sum.fill_(0)
        #self.random_gradient_sum.fill_(0)
       # print('ori ', self.ori_gradient_sum.view(-1, 1)[:10])
        #print("rec ", recovered_grad_in_cuda[0].view(-1,1)[:10])
        #print("recover time cost ", time2 - time1)
        return recovered_grad_in_cuda