-
Notifications
You must be signed in to change notification settings - Fork 2
/
architecture.txt
144 lines (143 loc) · 8.67 KB
/
architecture.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
This file contains architecture of default QuartzNet
(as described in the paper) applied to [-1, 128, 1080] input.
This scheme was generated by torchsummary.summary().
----------------------------------------------------------------
Layer (type) Output Shape Param #
================================================================
Conv1d-1 [-1, 128, 540] 4,352
Conv1d-2 [-1, 256, 540] 33,024
BatchNorm1d-3 [-1, 256, 540] 512
ReLU-4 [-1, 256, 540] 0
Conv1d-5 [-1, 256, 540] 8,704
Conv1d-6 [-1, 256, 540] 65,792
BatchNorm1d-7 [-1, 256, 540] 512
ReLU-8 [-1, 256, 540] 0
Conv1d-9 [-1, 256, 540] 8,704
Conv1d-10 [-1, 256, 540] 65,792
BatchNorm1d-11 [-1, 256, 540] 512
ReLU-12 [-1, 256, 540] 0
Conv1d-13 [-1, 256, 540] 8,704
Conv1d-14 [-1, 256, 540] 65,792
BatchNorm1d-15 [-1, 256, 540] 512
ReLU-16 [-1, 256, 540] 0
Conv1d-17 [-1, 256, 540] 8,704
Conv1d-18 [-1, 256, 540] 65,792
BatchNorm1d-19 [-1, 256, 540] 512
ReLU-20 [-1, 256, 540] 0
Conv1d-21 [-1, 256, 540] 8,704
Conv1d-22 [-1, 256, 540] 65,792
BatchNorm1d-23 [-1, 256, 540] 512
Conv1d-24 [-1, 256, 540] 65,792
BatchNorm1d-25 [-1, 256, 540] 512
ReLU-26 [-1, 256, 540] 0
BasicBlock-27 [-1, 256, 540] 0
Conv1d-28 [-1, 256, 540] 10,240
Conv1d-29 [-1, 256, 540] 65,792
BatchNorm1d-30 [-1, 256, 540] 512
ReLU-31 [-1, 256, 540] 0
Conv1d-32 [-1, 256, 540] 10,240
Conv1d-33 [-1, 256, 540] 65,792
BatchNorm1d-34 [-1, 256, 540] 512
ReLU-35 [-1, 256, 540] 0
Conv1d-36 [-1, 256, 540] 10,240
Conv1d-37 [-1, 256, 540] 65,792
BatchNorm1d-38 [-1, 256, 540] 512
ReLU-39 [-1, 256, 540] 0
Conv1d-40 [-1, 256, 540] 10,240
Conv1d-41 [-1, 256, 540] 65,792
BatchNorm1d-42 [-1, 256, 540] 512
ReLU-43 [-1, 256, 540] 0
Conv1d-44 [-1, 256, 540] 10,240
Conv1d-45 [-1, 256, 540] 65,792
BatchNorm1d-46 [-1, 256, 540] 512
Conv1d-47 [-1, 256, 540] 65,792
BatchNorm1d-48 [-1, 256, 540] 512
ReLU-49 [-1, 256, 540] 0
BasicBlock-50 [-1, 256, 540] 0
Conv1d-51 [-1, 256, 540] 13,312
Conv1d-52 [-1, 256, 540] 65,792
BatchNorm1d-53 [-1, 256, 540] 512
ReLU-54 [-1, 256, 540] 0
Conv1d-55 [-1, 256, 540] 13,312
Conv1d-56 [-1, 256, 540] 65,792
BatchNorm1d-57 [-1, 256, 540] 512
ReLU-58 [-1, 256, 540] 0
Conv1d-59 [-1, 256, 540] 13,312
Conv1d-60 [-1, 256, 540] 65,792
BatchNorm1d-61 [-1, 256, 540] 512
ReLU-62 [-1, 256, 540] 0
Conv1d-63 [-1, 256, 540] 13,312
Conv1d-64 [-1, 256, 540] 65,792
BatchNorm1d-65 [-1, 256, 540] 512
ReLU-66 [-1, 256, 540] 0
Conv1d-67 [-1, 256, 540] 13,312
Conv1d-68 [-1, 256, 540] 65,792
BatchNorm1d-69 [-1, 256, 540] 512
Conv1d-70 [-1, 256, 540] 65,792
BatchNorm1d-71 [-1, 256, 540] 512
ReLU-72 [-1, 256, 540] 0
BasicBlock-73 [-1, 256, 540] 0
Conv1d-74 [-1, 256, 540] 16,384
Conv1d-75 [-1, 512, 540] 131,584
BatchNorm1d-76 [-1, 512, 540] 1,024
ReLU-77 [-1, 512, 540] 0
Conv1d-78 [-1, 512, 540] 32,768
Conv1d-79 [-1, 512, 540] 262,656
BatchNorm1d-80 [-1, 512, 540] 1,024
ReLU-81 [-1, 512, 540] 0
Conv1d-82 [-1, 512, 540] 32,768
Conv1d-83 [-1, 512, 540] 262,656
BatchNorm1d-84 [-1, 512, 540] 1,024
ReLU-85 [-1, 512, 540] 0
Conv1d-86 [-1, 512, 540] 32,768
Conv1d-87 [-1, 512, 540] 262,656
BatchNorm1d-88 [-1, 512, 540] 1,024
ReLU-89 [-1, 512, 540] 0
Conv1d-90 [-1, 512, 540] 32,768
Conv1d-91 [-1, 512, 540] 262,656
BatchNorm1d-92 [-1, 512, 540] 1,024
Conv1d-93 [-1, 512, 540] 131,584
BatchNorm1d-94 [-1, 512, 540] 1,024
ReLU-95 [-1, 512, 540] 0
BasicBlock-96 [-1, 512, 540] 0
Conv1d-97 [-1, 512, 540] 38,912
Conv1d-98 [-1, 512, 540] 262,656
BatchNorm1d-99 [-1, 512, 540] 1,024
ReLU-100 [-1, 512, 540] 0
Conv1d-101 [-1, 512, 540] 38,912
Conv1d-102 [-1, 512, 540] 262,656
BatchNorm1d-103 [-1, 512, 540] 1,024
ReLU-104 [-1, 512, 540] 0
Conv1d-105 [-1, 512, 540] 38,912
Conv1d-106 [-1, 512, 540] 262,656
BatchNorm1d-107 [-1, 512, 540] 1,024
ReLU-108 [-1, 512, 540] 0
Conv1d-109 [-1, 512, 540] 38,912
Conv1d-110 [-1, 512, 540] 262,656
BatchNorm1d-111 [-1, 512, 540] 1,024
ReLU-112 [-1, 512, 540] 0
Conv1d-113 [-1, 512, 540] 38,912
Conv1d-114 [-1, 512, 540] 262,656
BatchNorm1d-115 [-1, 512, 540] 1,024
Conv1d-116 [-1, 512, 540] 262,656
BatchNorm1d-117 [-1, 512, 540] 1,024
ReLU-118 [-1, 512, 540] 0
BasicBlock-119 [-1, 512, 540] 0
Conv1d-120 [-1, 512, 540] 45,056
Conv1d-121 [-1, 512, 540] 262,656
BatchNorm1d-122 [-1, 512, 540] 1,024
ReLU-123 [-1, 512, 540] 0
Conv1d-124 [-1, 1024, 540] 525,312
BatchNorm1d-125 [-1, 1024, 540] 2,048
ReLU-126 [-1, 1024, 540] 0
Conv1d-127 [-1, 28, 540] 28,700
================================================================
Total params: 5,501,468
Trainable params: 5,501,468
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 0.53
Forward/backward pass size (MB): 193.65
Params size (MB): 20.99
Estimated Total Size (MB): 215.16
----------------------------------------------------------------