forked from MarcoForte/FBA_Matting
-
Notifications
You must be signed in to change notification settings - Fork 2
/
encoder-net.txt
187 lines (187 loc) · 11.4 KB
/
encoder-net.txt
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
Build Model
modifying input layer to accept 11 channels
----------------------------------------------------------------
Layer (type) Output Shape Param #
================================================================
Conv2d-1 [-1, 64, 160, 160] 34,496
GroupNorm-2 [-1, 64, 160, 160] 128
ReLU-3 [-1, 64, 160, 160] 0
MaxPool2d-4 [[-1, 64, 80, 80], [-1, 64, 80, 80]] 0
Conv2d-5 [-1, 64, 80, 80] 4,096
GroupNorm-6 [-1, 64, 80, 80] 128
ReLU-7 [-1, 64, 80, 80] 0
Conv2d-8 [-1, 64, 80, 80] 36,864
GroupNorm-9 [-1, 64, 80, 80] 128
ReLU-10 [-1, 64, 80, 80] 0
Conv2d-11 [-1, 256, 80, 80] 16,384
GroupNorm-12 [-1, 256, 80, 80] 512
Conv2d-13 [-1, 256, 80, 80] 16,384
GroupNorm-14 [-1, 256, 80, 80] 512
ReLU-15 [-1, 256, 80, 80] 0
Bottleneck-16 [-1, 256, 80, 80] 0
Conv2d-17 [-1, 64, 80, 80] 16,384
GroupNorm-18 [-1, 64, 80, 80] 128
ReLU-19 [-1, 64, 80, 80] 0
Conv2d-20 [-1, 64, 80, 80] 36,864
GroupNorm-21 [-1, 64, 80, 80] 128
ReLU-22 [-1, 64, 80, 80] 0
Conv2d-23 [-1, 256, 80, 80] 16,384
GroupNorm-24 [-1, 256, 80, 80] 512
ReLU-25 [-1, 256, 80, 80] 0
Bottleneck-26 [-1, 256, 80, 80] 0
Conv2d-27 [-1, 64, 80, 80] 16,384
GroupNorm-28 [-1, 64, 80, 80] 128
ReLU-29 [-1, 64, 80, 80] 0
Conv2d-30 [-1, 64, 80, 80] 36,864
GroupNorm-31 [-1, 64, 80, 80] 128
ReLU-32 [-1, 64, 80, 80] 0
Conv2d-33 [-1, 256, 80, 80] 16,384
GroupNorm-34 [-1, 256, 80, 80] 512
ReLU-35 [-1, 256, 80, 80] 0
Bottleneck-36 [-1, 256, 80, 80] 0
Conv2d-37 [-1, 128, 80, 80] 32,768
GroupNorm-38 [-1, 128, 80, 80] 256
ReLU-39 [-1, 128, 80, 80] 0
Conv2d-40 [-1, 128, 40, 40] 147,456
GroupNorm-41 [-1, 128, 40, 40] 256
ReLU-42 [-1, 128, 40, 40] 0
Conv2d-43 [-1, 512, 40, 40] 65,536
GroupNorm-44 [-1, 512, 40, 40] 1,024
Conv2d-45 [-1, 512, 40, 40] 131,072
GroupNorm-46 [-1, 512, 40, 40] 1,024
ReLU-47 [-1, 512, 40, 40] 0
Bottleneck-48 [-1, 512, 40, 40] 0
Conv2d-49 [-1, 128, 40, 40] 65,536
GroupNorm-50 [-1, 128, 40, 40] 256
ReLU-51 [-1, 128, 40, 40] 0
Conv2d-52 [-1, 128, 40, 40] 147,456
GroupNorm-53 [-1, 128, 40, 40] 256
ReLU-54 [-1, 128, 40, 40] 0
Conv2d-55 [-1, 512, 40, 40] 65,536
GroupNorm-56 [-1, 512, 40, 40] 1,024
ReLU-57 [-1, 512, 40, 40] 0
Bottleneck-58 [-1, 512, 40, 40] 0
Conv2d-59 [-1, 128, 40, 40] 65,536
GroupNorm-60 [-1, 128, 40, 40] 256
ReLU-61 [-1, 128, 40, 40] 0
Conv2d-62 [-1, 128, 40, 40] 147,456
GroupNorm-63 [-1, 128, 40, 40] 256
ReLU-64 [-1, 128, 40, 40] 0
Conv2d-65 [-1, 512, 40, 40] 65,536
GroupNorm-66 [-1, 512, 40, 40] 1,024
ReLU-67 [-1, 512, 40, 40] 0
Bottleneck-68 [-1, 512, 40, 40] 0
Conv2d-69 [-1, 128, 40, 40] 65,536
GroupNorm-70 [-1, 128, 40, 40] 256
ReLU-71 [-1, 128, 40, 40] 0
Conv2d-72 [-1, 128, 40, 40] 147,456
GroupNorm-73 [-1, 128, 40, 40] 256
ReLU-74 [-1, 128, 40, 40] 0
Conv2d-75 [-1, 512, 40, 40] 65,536
GroupNorm-76 [-1, 512, 40, 40] 1,024
ReLU-77 [-1, 512, 40, 40] 0
Bottleneck-78 [-1, 512, 40, 40] 0
Conv2d-79 [-1, 256, 40, 40] 131,072
GroupNorm-80 [-1, 256, 40, 40] 512
ReLU-81 [-1, 256, 40, 40] 0
Conv2d-82 [-1, 256, 40, 40] 589,824
GroupNorm-83 [-1, 256, 40, 40] 512
ReLU-84 [-1, 256, 40, 40] 0
Conv2d-85 [-1, 1024, 40, 40] 262,144
GroupNorm-86 [-1, 1024, 40, 40] 2,048
Conv2d-87 [-1, 1024, 40, 40] 524,288
GroupNorm-88 [-1, 1024, 40, 40] 2,048
ReLU-89 [-1, 1024, 40, 40] 0
Bottleneck-90 [-1, 1024, 40, 40] 0
Conv2d-91 [-1, 256, 40, 40] 262,144
GroupNorm-92 [-1, 256, 40, 40] 512
ReLU-93 [-1, 256, 40, 40] 0
Conv2d-94 [-1, 256, 40, 40] 589,824
GroupNorm-95 [-1, 256, 40, 40] 512
ReLU-96 [-1, 256, 40, 40] 0
Conv2d-97 [-1, 1024, 40, 40] 262,144
GroupNorm-98 [-1, 1024, 40, 40] 2,048
ReLU-99 [-1, 1024, 40, 40] 0
Bottleneck-100 [-1, 1024, 40, 40] 0
Conv2d-101 [-1, 256, 40, 40] 262,144
GroupNorm-102 [-1, 256, 40, 40] 512
ReLU-103 [-1, 256, 40, 40] 0
Conv2d-104 [-1, 256, 40, 40] 589,824
GroupNorm-105 [-1, 256, 40, 40] 512
ReLU-106 [-1, 256, 40, 40] 0
Conv2d-107 [-1, 1024, 40, 40] 262,144
GroupNorm-108 [-1, 1024, 40, 40] 2,048
ReLU-109 [-1, 1024, 40, 40] 0
Bottleneck-110 [-1, 1024, 40, 40] 0
Conv2d-111 [-1, 256, 40, 40] 262,144
GroupNorm-112 [-1, 256, 40, 40] 512
ReLU-113 [-1, 256, 40, 40] 0
Conv2d-114 [-1, 256, 40, 40] 589,824
GroupNorm-115 [-1, 256, 40, 40] 512
ReLU-116 [-1, 256, 40, 40] 0
Conv2d-117 [-1, 1024, 40, 40] 262,144
GroupNorm-118 [-1, 1024, 40, 40] 2,048
ReLU-119 [-1, 1024, 40, 40] 0
Bottleneck-120 [-1, 1024, 40, 40] 0
Conv2d-121 [-1, 256, 40, 40] 262,144
GroupNorm-122 [-1, 256, 40, 40] 512
ReLU-123 [-1, 256, 40, 40] 0
Conv2d-124 [-1, 256, 40, 40] 589,824
GroupNorm-125 [-1, 256, 40, 40] 512
ReLU-126 [-1, 256, 40, 40] 0
Conv2d-127 [-1, 1024, 40, 40] 262,144
GroupNorm-128 [-1, 1024, 40, 40] 2,048
ReLU-129 [-1, 1024, 40, 40] 0
Bottleneck-130 [-1, 1024, 40, 40] 0
Conv2d-131 [-1, 256, 40, 40] 262,144
GroupNorm-132 [-1, 256, 40, 40] 512
ReLU-133 [-1, 256, 40, 40] 0
Conv2d-134 [-1, 256, 40, 40] 589,824
GroupNorm-135 [-1, 256, 40, 40] 512
ReLU-136 [-1, 256, 40, 40] 0
Conv2d-137 [-1, 1024, 40, 40] 262,144
GroupNorm-138 [-1, 1024, 40, 40] 2,048
ReLU-139 [-1, 1024, 40, 40] 0
Bottleneck-140 [-1, 1024, 40, 40] 0
Conv2d-141 [-1, 512, 40, 40] 524,288
GroupNorm-142 [-1, 512, 40, 40] 1,024
ReLU-143 [-1, 512, 40, 40] 0
Conv2d-144 [-1, 512, 40, 40] 2,359,296
GroupNorm-145 [-1, 512, 40, 40] 1,024
ReLU-146 [-1, 512, 40, 40] 0
Conv2d-147 [-1, 2048, 40, 40] 1,048,576
GroupNorm-148 [-1, 2048, 40, 40] 4,096
Conv2d-149 [-1, 2048, 40, 40] 2,097,152
GroupNorm-150 [-1, 2048, 40, 40] 4,096
ReLU-151 [-1, 2048, 40, 40] 0
Bottleneck-152 [-1, 2048, 40, 40] 0
Conv2d-153 [-1, 512, 40, 40] 1,048,576
GroupNorm-154 [-1, 512, 40, 40] 1,024
ReLU-155 [-1, 512, 40, 40] 0
Conv2d-156 [-1, 512, 40, 40] 2,359,296
GroupNorm-157 [-1, 512, 40, 40] 1,024
ReLU-158 [-1, 512, 40, 40] 0
Conv2d-159 [-1, 2048, 40, 40] 1,048,576
GroupNorm-160 [-1, 2048, 40, 40] 4,096
ReLU-161 [-1, 2048, 40, 40] 0
Bottleneck-162 [-1, 2048, 40, 40] 0
Conv2d-163 [-1, 512, 40, 40] 1,048,576
GroupNorm-164 [-1, 512, 40, 40] 1,024
ReLU-165 [-1, 512, 40, 40] 0
Conv2d-166 [-1, 512, 40, 40] 2,359,296
GroupNorm-167 [-1, 512, 40, 40] 1,024
ReLU-168 [-1, 512, 40, 40] 0
Conv2d-169 [-1, 2048, 40, 40] 1,048,576
GroupNorm-170 [-1, 2048, 40, 40] 4,096
ReLU-171 [-1, 2048, 40, 40] 0
Bottleneck-172 [-1, 2048, 40, 40] 0
================================================================
Total params: 23,533,120
Trainable params: 23,533,120
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 4.30
Forward/backward pass size (MB): 1278667.19
Params size (MB): 89.77
Estimated Total Size (MB): 1278761.26
----------------------------------------------------------------