niobures commited on
Commit
7b6980a
·
verified ·
1 Parent(s): 461c6a8

Deep-Music-Enhancer (code, models, paper)

Browse files
Files changed (31) hide show
  1. .gitattributes +6 -0
  2. On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks.pdf +3 -0
  3. code/deep-music-enhancer [cringegaming64] +1 -2.zip +3 -0
  4. code/deep-music-enhancer.zip +3 -0
  5. models/ailia-models/code/LICENSE.md +648 -0
  6. models/ailia-models/code/README.md +85 -0
  7. models/ailia-models/code/deep_music_enhancer.py +174 -0
  8. models/ailia-models/code/deep_music_enhancer_utils.py +104 -0
  9. models/ailia-models/code/input.wav +3 -0
  10. models/ailia-models/code/input_butter_input_spec.png +0 -0
  11. models/ailia-models/code/input_butter_output_spec.png +3 -0
  12. models/ailia-models/code/input_cheby1_input_spec.png +3 -0
  13. models/ailia-models/code/input_cheby1_output_spec.png +3 -0
  14. models/ailia-models/code/output.wav +3 -0
  15. models/ailia-models/resnet.onnx +3 -0
  16. models/ailia-models/resnet.onnx.prototxt +2023 -0
  17. models/ailia-models/resnetbn.onnx +3 -0
  18. models/ailia-models/resnetbn.onnx.prototxt +2023 -0
  19. models/ailia-models/resnetda.onnx +3 -0
  20. models/ailia-models/resnetda.onnx.prototxt +2023 -0
  21. models/ailia-models/resnetdo.onnx +3 -0
  22. models/ailia-models/resnetdo.onnx.prototxt +2023 -0
  23. models/ailia-models/source.txt +25 -0
  24. models/ailia-models/unet.onnx +3 -0
  25. models/ailia-models/unet.onnx.prototxt +2028 -0
  26. models/ailia-models/unetbn.onnx +3 -0
  27. models/ailia-models/unetbn.onnx.prototxt +2028 -0
  28. models/ailia-models/unetda.onnx +3 -0
  29. models/ailia-models/unetda.onnx.prototxt +2028 -0
  30. models/ailia-models/unetdo.onnx +3 -0
  31. models/ailia-models/unetdo.onnx.prototxt +2028 -0
.gitattributes CHANGED
@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ models/ailia-models/code/input_butter_output_spec.png filter=lfs diff=lfs merge=lfs -text
37
+ models/ailia-models/code/input_cheby1_input_spec.png filter=lfs diff=lfs merge=lfs -text
38
+ models/ailia-models/code/input_cheby1_output_spec.png filter=lfs diff=lfs merge=lfs -text
39
+ models/ailia-models/code/input.wav filter=lfs diff=lfs merge=lfs -text
40
+ models/ailia-models/code/output.wav filter=lfs diff=lfs merge=lfs -text
41
+ On[[:space:]]Filter[[:space:]]Generalization[[:space:]]for[[:space:]]Music[[:space:]]Bandwidth[[:space:]]Extension[[:space:]]Using[[:space:]]Deep[[:space:]]Neural[[:space:]]Networks.pdf filter=lfs diff=lfs merge=lfs -text
On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6e62add03cb7802dac9d7168788d5c6aa4673f504d7097bd4694f26f27f7d3f
3
+ size 4503373
code/deep-music-enhancer [cringegaming64] +1 -2.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbd73c858a7e1622ed89e928509d0bf30a5177785181506d42c2092021291307
3
+ size 82495
code/deep-music-enhancer.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e6a5622d5eca91865af7c8af8c10aeb7cf5c2e4693d59c04211f54d8c83a770
3
+ size 82579
models/ailia-models/code/LICENSE.md ADDED
@@ -0,0 +1,648 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Copyright © 2019 INESC TEC 
2
+  
3
+ Deep music enhancer. Uses deep neural networks to extend the bandwidth of musical audio.
4
+
5
+ This software is authored by: 
6
+ Serkan Sulun
7
+
8
+
9
+ This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
10
+ This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
11
+ You should have received a copy of the GNU General Public License along with this program. If not, see <https://www.gnu.org/licenses/>.
12
+ You can reach INESC TEC at info@inesctec.pt, or
13
+ Campus da Faculdade de Engenharia da Universidade do Porto
14
+ Rua Dr. Roberto Frias
15
+ 4200-465 Porto
16
+ Portugal
17
+
18
+ A commercial license is also available for use in industrial projects and collaborations that do not wish to use the GPL 3 license.
19
+
20
+ If you use deep music enhancer in a work that leads to a scientific publication, we would appreciate it if you would kindly cite deep music enhancer in your manuscript.
21
+
22
+ S. Sulun and M. E. P. Davies, "On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks," in IEEE Journal of Selected Topics in Signal Processing
23
+
24
+ The paper can be found at https://doi.org/10.1109/JSTSP.2020.3037485
25
+
26
+
27
+
28
+
29
+
30
+ GNU GENERAL PUBLIC LICENSE
31
+ Version 3, 29 June 2007
32
+
33
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
34
+ Everyone is permitted to copy and distribute verbatim copies
35
+ of this license document, but changing it is not allowed.
36
+
37
+ Preamble
38
+
39
+ The GNU General Public License is a free, copyleft license for
40
+ software and other kinds of works.
41
+
42
+ The licenses for most software and other practical works are designed
43
+ to take away your freedom to share and change the works. By contrast,
44
+ the GNU General Public License is intended to guarantee your freedom to
45
+ share and change all versions of a program--to make sure it remains free
46
+ software for all its users. We, the Free Software Foundation, use the
47
+ GNU General Public License for most of our software; it applies also to
48
+ any other work released this way by its authors. You can apply it to
49
+ your programs, too.
50
+
51
+ When we speak of free software, we are referring to freedom, not
52
+ price. Our General Public Licenses are designed to make sure that you
53
+ have the freedom to distribute copies of free software (and charge for
54
+ them if you wish), that you receive source code or can get it if you
55
+ want it, that you can change the software or use pieces of it in new
56
+ free programs, and that you know you can do these things.
57
+
58
+ To protect your rights, we need to prevent others from denying you
59
+ these rights or asking you to surrender the rights. Therefore, you have
60
+ certain responsibilities if you distribute copies of the software, or if
61
+ you modify it: responsibilities to respect the freedom of others.
62
+
63
+ For example, if you distribute copies of such a program, whether
64
+ gratis or for a fee, you must pass on to the recipients the same
65
+ freedoms that you received. You must make sure that they, too, receive
66
+ or can get the source code. And you must show them these terms so they
67
+ know their rights.
68
+
69
+ Developers that use the GNU GPL protect your rights with two steps:
70
+ (1) assert copyright on the software, and (2) offer you this License
71
+ giving you legal permission to copy, distribute and/or modify it.
72
+
73
+ For the developers' and authors' protection, the GPL clearly explains
74
+ that there is no warranty for this free software. For both users' and
75
+ authors' sake, the GPL requires that modified versions be marked as
76
+ changed, so that their problems will not be attributed erroneously to
77
+ authors of previous versions.
78
+
79
+ Some devices are designed to deny users access to install or run
80
+ modified versions of the software inside them, although the manufacturer
81
+ can do so. This is fundamentally incompatible with the aim of
82
+ protecting users' freedom to change the software. The systematic
83
+ pattern of such abuse occurs in the area of products for individuals to
84
+ use, which is precisely where it is most unacceptable. Therefore, we
85
+ have designed this version of the GPL to prohibit the practice for those
86
+ products. If such problems arise substantially in other domains, we
87
+ stand ready to extend this provision to those domains in future versions
88
+ of the GPL, as needed to protect the freedom of users.
89
+
90
+ Finally, every program is threatened constantly by software patents.
91
+ States should not allow patents to restrict development and use of
92
+ software on general-purpose computers, but in those that do, we wish to
93
+ avoid the special danger that patents applied to a free program could
94
+ make it effectively proprietary. To prevent this, the GPL assures that
95
+ patents cannot be used to render the program non-free.
96
+
97
+ The precise terms and conditions for copying, distribution and
98
+ modification follow.
99
+
100
+ TERMS AND CONDITIONS
101
+
102
+ 0. Definitions.
103
+
104
+ "This License" refers to version 3 of the GNU General Public License.
105
+
106
+ "Copyright" also means copyright-like laws that apply to other kinds of
107
+ works, such as semiconductor masks.
108
+
109
+ "The Program" refers to any copyrightable work licensed under this
110
+ License. Each licensee is addressed as "you". "Licensees" and
111
+ "recipients" may be individuals or organizations.
112
+
113
+ To "modify" a work means to copy from or adapt all or part of the work
114
+ in a fashion requiring copyright permission, other than the making of an
115
+ exact copy. The resulting work is called a "modified version" of the
116
+ earlier work or a work "based on" the earlier work.
117
+
118
+ A "covered work" means either the unmodified Program or a work based
119
+ on the Program.
120
+
121
+ To "propagate" a work means to do anything with it that, without
122
+ permission, would make you directly or secondarily liable for
123
+ infringement under applicable copyright law, except executing it on a
124
+ computer or modifying a private copy. Propagation includes copying,
125
+ distribution (with or without modification), making available to the
126
+ public, and in some countries other activities as well.
127
+
128
+ To "convey" a work means any kind of propagation that enables other
129
+ parties to make or receive copies. Mere interaction with a user through
130
+ a computer network, with no transfer of a copy, is not conveying.
131
+
132
+ An interactive user interface displays "Appropriate Legal Notices"
133
+ to the extent that it includes a convenient and prominently visible
134
+ feature that (1) displays an appropriate copyright notice, and (2)
135
+ tells the user that there is no warranty for the work (except to the
136
+ extent that warranties are provided), that licensees may convey the
137
+ work under this License, and how to view a copy of this License. If
138
+ the interface presents a list of user commands or options, such as a
139
+ menu, a prominent item in the list meets this criterion.
140
+
141
+ 1. Source Code.
142
+
143
+ The "source code" for a work means the preferred form of the work
144
+ for making modifications to it. "Object code" means any non-source
145
+ form of a work.
146
+
147
+ A "Standard Interface" means an interface that either is an official
148
+ standard defined by a recognized standards body, or, in the case of
149
+ interfaces specified for a particular programming language, one that
150
+ is widely used among developers working in that language.
151
+
152
+ The "System Libraries" of an executable work include anything, other
153
+ than the work as a whole, that (a) is included in the normal form of
154
+ packaging a Major Component, but which is not part of that Major
155
+ Component, and (b) serves only to enable use of the work with that
156
+ Major Component, or to implement a Standard Interface for which an
157
+ implementation is available to the public in source code form. A
158
+ "Major Component", in this context, means a major essential component
159
+ (kernel, window system, and so on) of the specific operating system
160
+ (if any) on which the executable work runs, or a compiler used to
161
+ produce the work, or an object code interpreter used to run it.
162
+
163
+ The "Corresponding Source" for a work in object code form means all
164
+ the source code needed to generate, install, and (for an executable
165
+ work) run the object code and to modify the work, including scripts to
166
+ control those activities. However, it does not include the work's
167
+ System Libraries, or general-purpose tools or generally available free
168
+ programs which are used unmodified in performing those activities but
169
+ which are not part of the work. For example, Corresponding Source
170
+ includes interface definition files associated with source files for
171
+ the work, and the source code for shared libraries and dynamically
172
+ linked subprograms that the work is specifically designed to require,
173
+ such as by intimate data communication or control flow between those
174
+ subprograms and other parts of the work.
175
+
176
+ The Corresponding Source need not include anything that users
177
+ can regenerate automatically from other parts of the Corresponding
178
+ Source.
179
+
180
+ The Corresponding Source for a work in source code form is that
181
+ same work.
182
+
183
+ 2. Basic Permissions.
184
+
185
+ All rights granted under this License are granted for the term of
186
+ copyright on the Program, and are irrevocable provided the stated
187
+ conditions are met. This License explicitly affirms your unlimited
188
+ permission to run the unmodified Program. The output from running a
189
+ covered work is covered by this License only if the output, given its
190
+ content, constitutes a covered work. This License acknowledges your
191
+ rights of fair use or other equivalent, as provided by copyright law.
192
+
193
+ You may make, run and propagate covered works that you do not
194
+ convey, without conditions so long as your license otherwise remains
195
+ in force. You may convey covered works to others for the sole purpose
196
+ of having them make modifications exclusively for you, or provide you
197
+ with facilities for running those works, provided that you comply with
198
+ the terms of this License in conveying all material for which you do
199
+ not control copyright. Those thus making or running the covered works
200
+ for you must do so exclusively on your behalf, under your direction
201
+ and control, on terms that prohibit them from making any copies of
202
+ your copyrighted material outside their relationship with you.
203
+
204
+ Conveying under any other circumstances is permitted solely under
205
+ the conditions stated below. Sublicensing is not allowed; section 10
206
+ makes it unnecessary.
207
+
208
+ 3. Protecting Users' Legal Rights From Anti-Circumvention Law.
209
+
210
+ No covered work shall be deemed part of an effective technological
211
+ measure under any applicable law fulfilling obligations under article
212
+ 11 of the WIPO copyright treaty adopted on 20 December 1996, or
213
+ similar laws prohibiting or restricting circumvention of such
214
+ measures.
215
+
216
+ When you convey a covered work, you waive any legal power to forbid
217
+ circumvention of technological measures to the extent such circumvention
218
+ is effected by exercising rights under this License with respect to
219
+ the covered work, and you disclaim any intention to limit operation or
220
+ modification of the work as a means of enforcing, against the work's
221
+ users, your or third parties' legal rights to forbid circumvention of
222
+ technological measures.
223
+
224
+ 4. Conveying Verbatim Copies.
225
+
226
+ You may convey verbatim copies of the Program's source code as you
227
+ receive it, in any medium, provided that you conspicuously and
228
+ appropriately publish on each copy an appropriate copyright notice;
229
+ keep intact all notices stating that this License and any
230
+ non-permissive terms added in accord with section 7 apply to the code;
231
+ keep intact all notices of the absence of any warranty; and give all
232
+ recipients a copy of this License along with the Program.
233
+
234
+ You may charge any price or no price for each copy that you convey,
235
+ and you may offer support or warranty protection for a fee.
236
+
237
+ 5. Conveying Modified Source Versions.
238
+
239
+ You may convey a work based on the Program, or the modifications to
240
+ produce it from the Program, in the form of source code under the
241
+ terms of section 4, provided that you also meet all of these conditions:
242
+
243
+ a) The work must carry prominent notices stating that you modified
244
+ it, and giving a relevant date.
245
+
246
+ b) The work must carry prominent notices stating that it is
247
+ released under this License and any conditions added under section
248
+ 7. This requirement modifies the requirement in section 4 to
249
+ "keep intact all notices".
250
+
251
+ c) You must license the entire work, as a whole, under this
252
+ License to anyone who comes into possession of a copy. This
253
+ License will therefore apply, along with any applicable section 7
254
+ additional terms, to the whole of the work, and all its parts,
255
+ regardless of how they are packaged. This License gives no
256
+ permission to license the work in any other way, but it does not
257
+ invalidate such permission if you have separately received it.
258
+
259
+ d) If the work has interactive user interfaces, each must display
260
+ Appropriate Legal Notices; however, if the Program has interactive
261
+ interfaces that do not display Appropriate Legal Notices, your
262
+ work need not make them do so.
263
+
264
+ A compilation of a covered work with other separate and independent
265
+ works, which are not by their nature extensions of the covered work,
266
+ and which are not combined with it such as to form a larger program,
267
+ in or on a volume of a storage or distribution medium, is called an
268
+ "aggregate" if the compilation and its resulting copyright are not
269
+ used to limit the access or legal rights of the compilation's users
270
+ beyond what the individual works permit. Inclusion of a covered work
271
+ in an aggregate does not cause this License to apply to the other
272
+ parts of the aggregate.
273
+
274
+ 6. Conveying Non-Source Forms.
275
+
276
+ You may convey a covered work in object code form under the terms
277
+ of sections 4 and 5, provided that you also convey the
278
+ machine-readable Corresponding Source under the terms of this License,
279
+ in one of these ways:
280
+
281
+ a) Convey the object code in, or embodied in, a physical product
282
+ (including a physical distribution medium), accompanied by the
283
+ Corresponding Source fixed on a durable physical medium
284
+ customarily used for software interchange.
285
+
286
+ b) Convey the object code in, or embodied in, a physical product
287
+ (including a physical distribution medium), accompanied by a
288
+ written offer, valid for at least three years and valid for as
289
+ long as you offer spare parts or customer support for that product
290
+ model, to give anyone who possesses the object code either (1) a
291
+ copy of the Corresponding Source for all the software in the
292
+ product that is covered by this License, on a durable physical
293
+ medium customarily used for software interchange, for a price no
294
+ more than your reasonable cost of physically performing this
295
+ conveying of source, or (2) access to copy the
296
+ Corresponding Source from a network server at no charge.
297
+
298
+ c) Convey individual copies of the object code with a copy of the
299
+ written offer to provide the Corresponding Source. This
300
+ alternative is allowed only occasionally and noncommercially, and
301
+ only if you received the object code with such an offer, in accord
302
+ with subsection 6b.
303
+
304
+ d) Convey the object code by offering access from a designated
305
+ place (gratis or for a charge), and offer equivalent access to the
306
+ Corresponding Source in the same way through the same place at no
307
+ further charge. You need not require recipients to copy the
308
+ Corresponding Source along with the object code. If the place to
309
+ copy the object code is a network server, the Corresponding Source
310
+ may be on a different server (operated by you or a third party)
311
+ that supports equivalent copying facilities, provided you maintain
312
+ clear directions next to the object code saying where to find the
313
+ Corresponding Source. Regardless of what server hosts the
314
+ Corresponding Source, you remain obligated to ensure that it is
315
+ available for as long as needed to satisfy these requirements.
316
+
317
+ e) Convey the object code using peer-to-peer transmission, provided
318
+ you inform other peers where the object code and Corresponding
319
+ Source of the work are being offered to the general public at no
320
+ charge under subsection 6d.
321
+
322
+ A separable portion of the object code, whose source code is excluded
323
+ from the Corresponding Source as a System Library, need not be
324
+ included in conveying the object code work.
325
+
326
+ A "User Product" is either (1) a "consumer product", which means any
327
+ tangible personal property which is normally used for personal, family,
328
+ or household purposes, or (2) anything designed or sold for incorporation
329
+ into a dwelling. In determining whether a product is a consumer product,
330
+ doubtful cases shall be resolved in favor of coverage. For a particular
331
+ product received by a particular user, "normally used" refers to a
332
+ typical or common use of that class of product, regardless of the status
333
+ of the particular user or of the way in which the particular user
334
+ actually uses, or expects or is expected to use, the product. A product
335
+ is a consumer product regardless of whether the product has substantial
336
+ commercial, industrial or non-consumer uses, unless such uses represent
337
+ the only significant mode of use of the product.
338
+
339
+ "Installation Information" for a User Product means any methods,
340
+ procedures, authorization keys, or other information required to install
341
+ and execute modified versions of a covered work in that User Product from
342
+ a modified version of its Corresponding Source. The information must
343
+ suffice to ensure that the continued functioning of the modified object
344
+ code is in no case prevented or interfered with solely because
345
+ modification has been made.
346
+
347
+ If you convey an object code work under this section in, or with, or
348
+ specifically for use in, a User Product, and the conveying occurs as
349
+ part of a transaction in which the right of possession and use of the
350
+ User Product is transferred to the recipient in perpetuity or for a
351
+ fixed term (regardless of how the transaction is characterized), the
352
+ Corresponding Source conveyed under this section must be accompanied
353
+ by the Installation Information. But this requirement does not apply
354
+ if neither you nor any third party retains the ability to install
355
+ modified object code on the User Product (for example, the work has
356
+ been installed in ROM).
357
+
358
+ The requirement to provide Installation Information does not include a
359
+ requirement to continue to provide support service, warranty, or updates
360
+ for a work that has been modified or installed by the recipient, or for
361
+ the User Product in which it has been modified or installed. Access to a
362
+ network may be denied when the modification itself materially and
363
+ adversely affects the operation of the network or violates the rules and
364
+ protocols for communication across the network.
365
+
366
+ Corresponding Source conveyed, and Installation Information provided,
367
+ in accord with this section must be in a format that is publicly
368
+ documented (and with an implementation available to the public in
369
+ source code form), and must require no special password or key for
370
+ unpacking, reading or copying.
371
+
372
+ 7. Additional Terms.
373
+
374
+ "Additional permissions" are terms that supplement the terms of this
375
+ License by making exceptions from one or more of its conditions.
376
+ Additional permissions that are applicable to the entire Program shall
377
+ be treated as though they were included in this License, to the extent
378
+ that they are valid under applicable law. If additional permissions
379
+ apply only to part of the Program, that part may be used separately
380
+ under those permissions, but the entire Program remains governed by
381
+ this License without regard to the additional permissions.
382
+
383
+ When you convey a copy of a covered work, you may at your option
384
+ remove any additional permissions from that copy, or from any part of
385
+ it. (Additional permissions may be written to require their own
386
+ removal in certain cases when you modify the work.) You may place
387
+ additional permissions on material, added by you to a covered work,
388
+ for which you have or can give appropriate copyright permission.
389
+
390
+ Notwithstanding any other provision of this License, for material you
391
+ add to a covered work, you may (if authorized by the copyright holders of
392
+ that material) supplement the terms of this License with terms:
393
+
394
+ a) Disclaiming warranty or limiting liability differently from the
395
+ terms of sections 15 and 16 of this License; or
396
+
397
+ b) Requiring preservation of specified reasonable legal notices or
398
+ author attributions in that material or in the Appropriate Legal
399
+ Notices displayed by works containing it; or
400
+
401
+ c) Prohibiting misrepresentation of the origin of that material, or
402
+ requiring that modified versions of such material be marked in
403
+ reasonable ways as different from the original version; or
404
+
405
+ d) Limiting the use for publicity purposes of names of licensors or
406
+ authors of the material; or
407
+
408
+ e) Declining to grant rights under trademark law for use of some
409
+ trade names, trademarks, or service marks; or
410
+
411
+ f) Requiring indemnification of licensors and authors of that
412
+ material by anyone who conveys the material (or modified versions of
413
+ it) with contractual assumptions of liability to the recipient, for
414
+ any liability that these contractual assumptions directly impose on
415
+ those licensors and authors.
416
+
417
+ All other non-permissive additional terms are considered "further
418
+ restrictions" within the meaning of section 10. If the Program as you
419
+ received it, or any part of it, contains a notice stating that it is
420
+ governed by this License along with a term that is a further
421
+ restriction, you may remove that term. If a license document contains
422
+ a further restriction but permits relicensing or conveying under this
423
+ License, you may add to a covered work material governed by the terms
424
+ of that license document, provided that the further restriction does
425
+ not survive such relicensing or conveying.
426
+
427
+ If you add terms to a covered work in accord with this section, you
428
+ must place, in the relevant source files, a statement of the
429
+ additional terms that apply to those files, or a notice indicating
430
+ where to find the applicable terms.
431
+
432
+ Additional terms, permissive or non-permissive, may be stated in the
433
+ form of a separately written license, or stated as exceptions;
434
+ the above requirements apply either way.
435
+
436
+ 8. Termination.
437
+
438
+ You may not propagate or modify a covered work except as expressly
439
+ provided under this License. Any attempt otherwise to propagate or
440
+ modify it is void, and will automatically terminate your rights under
441
+ this License (including any patent licenses granted under the third
442
+ paragraph of section 11).
443
+
444
+ However, if you cease all violation of this License, then your
445
+ license from a particular copyright holder is reinstated (a)
446
+ provisionally, unless and until the copyright holder explicitly and
447
+ finally terminates your license, and (b) permanently, if the copyright
448
+ holder fails to notify you of the violation by some reasonable means
449
+ prior to 60 days after the cessation.
450
+
451
+ Moreover, your license from a particular copyright holder is
452
+ reinstated permanently if the copyright holder notifies you of the
453
+ violation by some reasonable means, this is the first time you have
454
+ received notice of violation of this License (for any work) from that
455
+ copyright holder, and you cure the violation prior to 30 days after
456
+ your receipt of the notice.
457
+
458
+ Termination of your rights under this section does not terminate the
459
+ licenses of parties who have received copies or rights from you under
460
+ this License. If your rights have been terminated and not permanently
461
+ reinstated, you do not qualify to receive new licenses for the same
462
+ material under section 10.
463
+
464
+ 9. Acceptance Not Required for Having Copies.
465
+
466
+ You are not required to accept this License in order to receive or
467
+ run a copy of the Program. Ancillary propagation of a covered work
468
+ occurring solely as a consequence of using peer-to-peer transmission
469
+ to receive a copy likewise does not require acceptance. However,
470
+ nothing other than this License grants you permission to propagate or
471
+ modify any covered work. These actions infringe copyright if you do
472
+ not accept this License. Therefore, by modifying or propagating a
473
+ covered work, you indicate your acceptance of this License to do so.
474
+
475
+ 10. Automatic Licensing of Downstream Recipients.
476
+
477
+ Each time you convey a covered work, the recipient automatically
478
+ receives a license from the original licensors, to run, modify and
479
+ propagate that work, subject to this License. You are not responsible
480
+ for enforcing compliance by third parties with this License.
481
+
482
+ An "entity transaction" is a transaction transferring control of an
483
+ organization, or substantially all assets of one, or subdividing an
484
+ organization, or merging organizations. If propagation of a covered
485
+ work results from an entity transaction, each party to that
486
+ transaction who receives a copy of the work also receives whatever
487
+ licenses to the work the party's predecessor in interest had or could
488
+ give under the previous paragraph, plus a right to possession of the
489
+ Corresponding Source of the work from the predecessor in interest, if
490
+ the predecessor has it or can get it with reasonable efforts.
491
+
492
+ You may not impose any further restrictions on the exercise of the
493
+ rights granted or affirmed under this License. For example, you may
494
+ not impose a license fee, royalty, or other charge for exercise of
495
+ rights granted under this License, and you may not initiate litigation
496
+ (including a cross-claim or counterclaim in a lawsuit) alleging that
497
+ any patent claim is infringed by making, using, selling, offering for
498
+ sale, or importing the Program or any portion of it.
499
+
500
+ 11. Patents.
501
+
502
+ A "contributor" is a copyright holder who authorizes use under this
503
+ License of the Program or a work on which the Program is based. The
504
+ work thus licensed is called the contributor's "contributor version".
505
+
506
+ A contributor's "essential patent claims" are all patent claims
507
+ owned or controlled by the contributor, whether already acquired or
508
+ hereafter acquired, that would be infringed by some manner, permitted
509
+ by this License, of making, using, or selling its contributor version,
510
+ but do not include claims that would be infringed only as a
511
+ consequence of further modification of the contributor version. For
512
+ purposes of this definition, "control" includes the right to grant
513
+ patent sublicenses in a manner consistent with the requirements of
514
+ this License.
515
+
516
+ Each contributor grants you a non-exclusive, worldwide, royalty-free
517
+ patent license under the contributor's essential patent claims, to
518
+ make, use, sell, offer for sale, import and otherwise run, modify and
519
+ propagate the contents of its contributor version.
520
+
521
+ In the following three paragraphs, a "patent license" is any express
522
+ agreement or commitment, however denominated, not to enforce a patent
523
+ (such as an express permission to practice a patent or covenant not to
524
+ sue for patent infringement). To "grant" such a patent license to a
525
+ party means to make such an agreement or commitment not to enforce a
526
+ patent against the party.
527
+
528
+ If you convey a covered work, knowingly relying on a patent license,
529
+ and the Corresponding Source of the work is not available for anyone
530
+ to copy, free of charge and under the terms of this License, through a
531
+ publicly available network server or other readily accessible means,
532
+ then you must either (1) cause the Corresponding Source to be so
533
+ available, or (2) arrange to deprive yourself of the benefit of the
534
+ patent license for this particular work, or (3) arrange, in a manner
535
+ consistent with the requirements of this License, to extend the patent
536
+ license to downstream recipients. "Knowingly relying" means you have
537
+ actual knowledge that, but for the patent license, your conveying the
538
+ covered work in a country, or your recipient's use of the covered work
539
+ in a country, would infringe one or more identifiable patents in that
540
+ country that you have reason to believe are valid.
541
+
542
+ If, pursuant to or in connection with a single transaction or
543
+ arrangement, you convey, or propagate by procuring conveyance of, a
544
+ covered work, and grant a patent license to some of the parties
545
+ receiving the covered work authorizing them to use, propagate, modify
546
+ or convey a specific copy of the covered work, then the patent license
547
+ you grant is automatically extended to all recipients of the covered
548
+ work and works based on it.
549
+
550
+ A patent license is "discriminatory" if it does not include within
551
+ the scope of its coverage, prohibits the exercise of, or is
552
+ conditioned on the non-exercise of one or more of the rights that are
553
+ specifically granted under this License. You may not convey a covered
554
+ work if you are a party to an arrangement with a third party that is
555
+ in the business of distributing software, under which you make payment
556
+ to the third party based on the extent of your activity of conveying
557
+ the work, and under which the third party grants, to any of the
558
+ parties who would receive the covered work from you, a discriminatory
559
+ patent license (a) in connection with copies of the covered work
560
+ conveyed by you (or copies made from those copies), or (b) primarily
561
+ for and in connection with specific products or compilations that
562
+ contain the covered work, unless you entered into that arrangement,
563
+ or that patent license was granted, prior to 28 March 2007.
564
+
565
+ Nothing in this License shall be construed as excluding or limiting
566
+ any implied license or other defenses to infringement that may
567
+ otherwise be available to you under applicable patent law.
568
+
569
+ 12. No Surrender of Others' Freedom.
570
+
571
+ If conditions are imposed on you (whether by court order, agreement or
572
+ otherwise) that contradict the conditions of this License, they do not
573
+ excuse you from the conditions of this License. If you cannot convey a
574
+ covered work so as to satisfy simultaneously your obligations under this
575
+ License and any other pertinent obligations, then as a consequence you may
576
+ not convey it at all. For example, if you agree to terms that obligate you
577
+ to collect a royalty for further conveying from those to whom you convey
578
+ the Program, the only way you could satisfy both those terms and this
579
+ License would be to refrain entirely from conveying the Program.
580
+
581
+ 13. Use with the GNU Affero General Public License.
582
+
583
+ Notwithstanding any other provision of this License, you have
584
+ permission to link or combine any covered work with a work licensed
585
+ under version 3 of the GNU Affero General Public License into a single
586
+ combined work, and to convey the resulting work. The terms of this
587
+ License will continue to apply to the part which is the covered work,
588
+ but the special requirements of the GNU Affero General Public License,
589
+ section 13, concerning interaction through a network will apply to the
590
+ combination as such.
591
+
592
+ 14. Revised Versions of this License.
593
+
594
+ The Free Software Foundation may publish revised and/or new versions of
595
+ the GNU General Public License from time to time. Such new versions will
596
+ be similar in spirit to the present version, but may differ in detail to
597
+ address new problems or concerns.
598
+
599
+ Each version is given a distinguishing version number. If the
600
+ Program specifies that a certain numbered version of the GNU General
601
+ Public License "or any later version" applies to it, you have the
602
+ option of following the terms and conditions either of that numbered
603
+ version or of any later version published by the Free Software
604
+ Foundation. If the Program does not specify a version number of the
605
+ GNU General Public License, you may choose any version ever published
606
+ by the Free Software Foundation.
607
+
608
+ If the Program specifies that a proxy can decide which future
609
+ versions of the GNU General Public License can be used, that proxy's
610
+ public statement of acceptance of a version permanently authorizes you
611
+ to choose that version for the Program.
612
+
613
+ Later license versions may give you additional or different
614
+ permissions. However, no additional obligations are imposed on any
615
+ author or copyright holder as a result of your choosing to follow a
616
+ later version.
617
+
618
+ 15. Disclaimer of Warranty.
619
+
620
+ THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
621
+ APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
622
+ HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
623
+ OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
624
+ THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
625
+ PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
626
+ IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
627
+ ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
628
+
629
+ 16. Limitation of Liability.
630
+
631
+ IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
632
+ WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
633
+ THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
634
+ GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
635
+ USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
636
+ DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
637
+ PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
638
+ EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
639
+ SUCH DAMAGES.
640
+
641
+ 17. Interpretation of Sections 15 and 16.
642
+
643
+ If the disclaimer of warranty and limitation of liability provided
644
+ above cannot be given local legal effect according to their terms,
645
+ reviewing courts shall apply local law that most closely approximates
646
+ an absolute waiver of all civil liability in connection with the
647
+ Program, unless a warranty or assumption of liability accompanies a
648
+ copy of the Program in return for a fee.
models/ailia-models/code/README.md ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks
2
+
3
+ ## Input
4
+
5
+ Audio file (.wav file)
6
+
7
+ input.wav is `(/Test/003 - Actions - One Minute Smile/mixture.wav)` in DSD100 dataset. (can be donwloaded from http://liutkus.net/DSD100.zip)
8
+ To reduce calculation cost, input.wav is clipped from original.
9
+
10
+
11
+ ## Output
12
+
13
+ Bandwidth extented audio file (.wav file)
14
+
15
+
16
+ ## Usage
17
+ Automatically downloads the onnx and prototxt files on the first run.
18
+ It is necessary to be connected to the Internet while downloading.
19
+
20
+ For the sample wav,
21
+ ```bash
22
+ $ python3 deep_music_enhancer.py
23
+ ```
24
+
25
+ Supported model types are [`resnet`, `resnet_bn`, `resnet_da`, `resnet_do`, `unet`, `unet_bn`, `unet_da`, `unet_do`].
26
+ bn means batch normlization, do means dropout, da means data augmentation.
27
+ Model type can be specified as below.
28
+ ```
29
+ $ python3 deep_music_enhancer.py --model [MODEL TYPE]
30
+ ```
31
+
32
+
33
+ You can specify input audio files by adding `--input` option.
34
+
35
+ ```
36
+ $ python3 deep_music_enhancer.py --input [INPUT WAV FILE]
37
+ ```
38
+
39
+ If you save audio output with specified name, you have to add `--savefile` option.
40
+
41
+ ```
42
+ $ python3 deep_music_enhancer.py --savepath [OUTPUT NAME]
43
+ ```
44
+
45
+ Additionaly, you can use `--vis` option in order to visualize spectrogram of input and output audio.
46
+
47
+ Spectrogram of input audio
48
+ ![Spectrogram of input audio](input_butter_input_spec.png "Spectrogram of input audio")
49
+
50
+ Spectrogram of output audio (butter filter)
51
+ ![Spectrogram of output audio (butter filter)](input_butter_output_spec.png "Spectrogram of output audio (butter filter)")
52
+
53
+ Spectrogram of output audio (cheby1 filter)
54
+ ![Spectrogram of output audio (cheby1 filter)](input_cheby1_output_spec.png "Spectrogram of output audio (cheby1 filter)")
55
+
56
+
57
+ ## Reference
58
+
59
+ [deep-music-enhancer](https://github.com/serkansulun/deep-music-enhancer)
60
+
61
+
62
+ ## Framework
63
+
64
+ Pytorch
65
+
66
+ ## Model Format
67
+
68
+ ONNX opset=11
69
+
70
+ ## Netron
71
+ [resnet.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnet.onnx.prototxt)
72
+
73
+ [resnetbn.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetbn.onnx.prototxt)
74
+
75
+ [resnetda.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetda.onnx.prototxt)
76
+
77
+ [resnetdo.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetdo.onnx.prototxt)
78
+
79
+ [unet.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/unet.onnx.prototxt)
80
+
81
+ [unetbn.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetbn.onnx.prototxt)
82
+
83
+ [unetda.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetda.onnx.prototxt)
84
+
85
+ [unetdo.onnx.prototxt](https://netron.app/?url=https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetdo.onnx.prototxt)
models/ailia-models/code/deep_music_enhancer.py ADDED
@@ -0,0 +1,174 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import time
2
+ import sys
3
+ import argparse
4
+
5
+ import numpy as np
6
+
7
+ import ailia # noqa: E402
8
+
9
+ # import original modules
10
+ sys.path.append('../../util')
11
+ from arg_utils import get_base_parser, update_parser, get_savepath # noqa: E402
12
+ from model_utils import check_and_download_models # noqa: E402
13
+
14
+ # logger
15
+ from logging import getLogger # noqa: E402
16
+ logger = getLogger(__name__)
17
+
18
+ import os
19
+ from tqdm import tqdm
20
+ import matplotlib.pyplot as plt
21
+ from scipy.io import wavfile
22
+ from deep_music_enhancer_utils import (
23
+ read_audio,
24
+ SingleSong
25
+ )
26
+
27
+
28
+ # ======================
29
+ # PARAMETERS
30
+ # ======================
31
+ WAV_PATH = 'input.wav'
32
+ SAVE_WAV_PATH = 'output.wav'
33
+
34
+ WEIGHT_PATH_RESNET = 'resnet.onnx'
35
+ MODEL_PATH_RESNET = 'resnet.onnx.prototxt'
36
+ WEIGHT_PATH_RESNET_BN = 'resnetbn.onnx'
37
+ MODEL_PATH_RESNET_BN = 'resnetbn.onnx.prototxt'
38
+ WEIGHT_PATH_RESNET_DA = 'resnetda.onnx'
39
+ MODEL_PATH_RESNET_DA = 'resnetda.onnx.prototxt'
40
+ WEIGHT_PATH_RESNET_DO = 'resnetdo.onnx'
41
+ MODEL_PATH_RESNET_DO = 'resnetdo.onnx.prototxt'
42
+
43
+ WEIGHT_PATH_UNET = 'unet.onnx'
44
+ MODEL_PATH_UNET = 'unet.onnx.prototxt'
45
+ WEIGHT_PATH_UNET_BN = 'unetbn.onnx'
46
+ MODEL_PATH_UNET_BN = 'unetbn.onnx.prototxt'
47
+ WEIGHT_PATH_UNET_DA = 'unetda.onnx'
48
+ MODEL_PATH_UNET_DA = 'unetda.onnx.prototxt'
49
+ WEIGHT_PATH_UNET_DO = 'unetdo.onnx'
50
+ MODEL_PATH_UNET_DO = 'unetdo.onnx.prototxt'
51
+
52
+ REMOTE_PATH = 'https://storage.googleapis.com/ailia-models/deep-music-enhancer/'
53
+
54
+
55
+ # ======================
56
+ # Arguemnt Parser Config
57
+ # ======================
58
+ parser = get_base_parser(
59
+ 'On Filter Generalization for Music Bandwidth Extension Using Deep Neural Networks',
60
+ WAV_PATH,
61
+ SAVE_WAV_PATH
62
+ )
63
+ # overwrite
64
+ parser.add_argument(
65
+ '--input', '-i', metavar='WAV', default=WAV_PATH,
66
+ help='input audio'
67
+ )
68
+ parser.add_argument(
69
+ '--ailia_audio', action='store_true',
70
+ help='use ailia audio library'
71
+ )
72
+ parser.add_argument(
73
+ '--vis', action='store_true',
74
+ help='save visualized spectrogram'
75
+ )
76
+ parser.add_argument(
77
+ '--model', type=str, default='unet',
78
+ choices=[
79
+ 'resnet', 'resnet_bn', 'resnet_da', 'resnet_do',
80
+ 'unet', 'unet_bn', 'unet_da', 'unet_do'
81
+ ],
82
+ )
83
+ args = update_parser(parser, check_input_type=False)
84
+
85
+
86
+ # ======================
87
+ # Main function
88
+ # ======================
89
+ def audio_bandwidth_extension(net):
90
+ FILTERS_TEST = [('cheby1', 6), ('butter', 6)]
91
+ c_SAMPLE_RATE = 44100
92
+ c_WAV_SAMPLE_LEN = 8192
93
+ cutoff = 11025
94
+ duration = None
95
+ start = 0
96
+
97
+ for filter_ in FILTERS_TEST:
98
+ input_name = args.input[0]
99
+ input_name_without_ext = os.path.splitext(os.path.basename(input_name))[0]
100
+ hq_path = input_name
101
+
102
+ logger.info('filter: {}, input_name: {}'.format(filter_, input_name))
103
+
104
+ # create dataset
105
+ song_data = SingleSong(
106
+ c_WAV_SAMPLE_LEN,
107
+ filter_,
108
+ hq_path,
109
+ cutoff=cutoff,
110
+ duration=duration,
111
+ start=start
112
+ )
113
+
114
+ y_full = song_data.preallocate() # preallocation to keep individual output chunks
115
+
116
+ idx_start_chunk = 0 # model works on chunks of audio, these are concatenated later
117
+
118
+ for i in tqdm(range(len(song_data))):
119
+ x, t = song_data[i]
120
+ x = x[np.newaxis, :, :]
121
+
122
+ y = net.predict(x)
123
+
124
+ idx_end_chunk = idx_start_chunk + y.shape[0]
125
+ y_full[idx_start_chunk:idx_end_chunk] = y
126
+ idx_start_chunk = idx_end_chunk
127
+
128
+ y_full = np.concatenate(y_full, axis=-1) # create full song out of chunks
129
+
130
+ x_full, t_full = song_data.get_full_signals()
131
+ y_full = np.clip(y_full, -1, 1 - np.finfo(np.float32).eps)
132
+
133
+ # save audio
134
+ wavfile.write(args.savepath, c_SAMPLE_RATE, y_full.T)
135
+
136
+ # save spec
137
+ if args.vis:
138
+ _, _, _, _ = plt.specgram(x_full.T[:c_SAMPLE_RATE*5, 0], Fs=c_SAMPLE_RATE)
139
+ plt.savefig('{}_{}_input_spec.png'.format(input_name_without_ext, filter_[0]))
140
+ _, _, _, _ = plt.specgram(y_full.T[:c_SAMPLE_RATE*5, 0], Fs=c_SAMPLE_RATE)
141
+ plt.savefig('{}_{}_output_spec.png'.format(input_name_without_ext, filter_[0]))
142
+
143
+ logger.info('Script finished successfully.')
144
+
145
+
146
+ def main():
147
+ # model files check and download
148
+ if args.model == 'resnet':
149
+ weight_path, model_path = WEIGHT_PATH_RESNET, MODEL_PATH_RESNET
150
+ elif args.model == 'resnet_bn':
151
+ weight_path, model_path = WEIGHT_PATH_RESNET_BN, MODEL_PATH_RESNET_BN
152
+ elif args.model == 'resnet_da':
153
+ weight_path, model_path = WEIGHT_PATH_RESNET_DA, MODEL_PATH_RESNET_DA
154
+ elif args.model == 'resnet_do':
155
+ weight_path, model_path = WEIGHT_PATH_RESNET_DO, MODEL_PATH_RESNET_DO
156
+ elif args.model == 'unet':
157
+ weight_path, model_path = WEIGHT_PATH_UNET, MODEL_PATH_UNET
158
+ elif args.model == 'unet_bn':
159
+ weight_path, model_path = WEIGHT_PATH_UNET_BN, MODEL_PATH_UNET_BN
160
+ elif args.model == 'unet_da':
161
+ weight_path, model_path = WEIGHT_PATH_UNET_DA, MODEL_PATH_UNET_DA
162
+ elif args.model == 'unet_do':
163
+ weight_path, model_path = WEIGHT_PATH_UNET_DO, MODEL_PATH_UNET_DO
164
+
165
+ env_id = args.env_id
166
+
167
+ check_and_download_models(weight_path, model_path, REMOTE_PATH)
168
+ net = ailia.Net(model_path, weight_path, env_id=env_id)
169
+
170
+ audio_bandwidth_extension(net)
171
+
172
+
173
+ if __name__ == "__main__":
174
+ main()
models/ailia-models/code/deep_music_enhancer_utils.py ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from scipy.io import wavfile
2
+ from scipy import signal
3
+ import numpy as np
4
+ import matplotlib.pyplot as plt
5
+
6
+
7
+ class SingleSong:
8
+ # To load one excerpt with arbitrary length, or one full song, for test or validation
9
+ def __init__(self, chunk_len, filter_, hq_path, cutoff, duration=None, start=8):
10
+
11
+ hq, sr = read_audio(hq_path) # high quality target
12
+ lq = lowpass(hq, cutoff, filter_=filter_) # low quality input
13
+
14
+ # CROP
15
+ song_len = lq.shape[-1]
16
+
17
+ if duration is None: # save entire song
18
+ test_start = 0
19
+ test_len = song_len
20
+ else:
21
+ test_start = start * sr # start from n th second
22
+ test_len = duration * sr
23
+
24
+ test_len = min(test_len, song_len - test_start)
25
+
26
+ lq = lq[:, test_start:test_start + test_len]
27
+ hq = hq[:, test_start:test_start + test_len]
28
+
29
+ self.x_full = lq.copy()
30
+ self.t_full = hq.copy()
31
+
32
+ # To have equal length chunks for minibatching
33
+ time_len = lq.shape[-1]
34
+ n_chunks, rem = divmod(time_len, chunk_len)
35
+ lq = lq[..., :-rem or None] # or None handles rem=0
36
+ hq = hq[..., :-rem or None]
37
+
38
+ # adjust lengths
39
+ self.x_full = self.x_full[..., :lq.shape[-1] or None]
40
+ self.t_full = self.t_full[..., :lq.shape[-1] or None]
41
+
42
+ # Save full samples
43
+
44
+ self.lq = np.split(lq, n_chunks, axis=-1) # create a lists of chunks
45
+ self.hq = np.split(hq, n_chunks, axis=-1) # create a lists of chunks
46
+
47
+ def get_full_signals(self):
48
+ # Returns full length input and target
49
+ return self.x_full, self.t_full
50
+
51
+ def preallocate(self):
52
+ """
53
+ Preallocates the matrix to save all minibatch outputs.
54
+ It is faster to transfer all minibatches from GPU to CPU at once.
55
+ """
56
+ return np.zeros((len(self.lq), *self.lq[0].shape))
57
+
58
+ def __len__(self):
59
+ return len(self.lq)
60
+
61
+ def __getitem__(self, idx):
62
+ return self.lq[idx], self.hq[idx]
63
+
64
+
65
+ def lowpass(sig, cutoff, filter_=('cheby1', 8), sr=44100):
66
+ """Lowpasses input signal based on a cutoff frequency
67
+
68
+ Arguments:
69
+ sig {numpy 1d array} -- input signal
70
+ cutoff {int} -- cutoff frequency
71
+
72
+ Keyword Arguments:
73
+ sr {int} -- sampling rate of the input signal (default: {44100})
74
+ filter_type {str} -- type of filter, only butter and cheby1 are implemented (default: {'butter'})
75
+
76
+ Returns:
77
+ numpy 1d array -- lowpassed signal
78
+ """
79
+ nyq = sr / 2
80
+ cutoff /= nyq
81
+
82
+ if filter_[0] == 'butter':
83
+ B, A = signal.butter(filter_[1], cutoff)
84
+ elif filter_[0] == 'cheby1':
85
+ B, A = signal.cheby1(filter_[1], 0.05, cutoff)
86
+ elif filter_[0] == 'bessel':
87
+ B, A = signal.bessel(filter_[1], cutoff, norm='mag')
88
+ elif filter_[0] == 'ellip':
89
+ B, A = signal.ellip(filter_[1], 0.05, 20, cutoff)
90
+
91
+ sig_lp = signal.filtfilt(B, A, sig)
92
+ return sig_lp.astype(np.float32)
93
+
94
+
95
+ def read_audio(path, make_stereo=True):
96
+ sr, audio = wavfile.read(path)
97
+ audio = audio.T
98
+ if np.issubdtype(audio.dtype, np.int16):
99
+ audio = audio.astype(np.float32) / 32768.0
100
+ if len(audio.shape) == 1: # if mono
101
+ audio = np.expand_dims(audio, axis=0)
102
+ if make_stereo:
103
+ audio = np.repeat(audio, 2, axis=0)
104
+ return audio, sr
models/ailia-models/code/input.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb101ea651efe0b97d554ab01ce85bf7ff7919f0a989a0b7b3b0d0d82f91f6dd
3
+ size 2721504
models/ailia-models/code/input_butter_input_spec.png ADDED
models/ailia-models/code/input_butter_output_spec.png ADDED

Git LFS Details

  • SHA256: bb070ba35fac5df6d388208db77cd735d9a5e74ac4c952b795a656f0fd53d164
  • Pointer size: 131 Bytes
  • Size of remote file: 202 kB
models/ailia-models/code/input_cheby1_input_spec.png ADDED

Git LFS Details

  • SHA256: 4efe657b92b69cef65b4ceb65d3051f393d74e76077ac93c8e46ac68f8ee343a
  • Pointer size: 131 Bytes
  • Size of remote file: 116 kB
models/ailia-models/code/input_cheby1_output_spec.png ADDED

Git LFS Details

  • SHA256: 27eb833aa348ddcf581fb6e0fd768fe52d0c7e011fad8e42f3ee5d65bf493b57
  • Pointer size: 131 Bytes
  • Size of remote file: 202 kB
models/ailia-models/code/output.wav ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf30b706e6182e359cd5411597c1ebd7a6f126b194d68910f2851c0cd01459b7
3
+ size 10747962
models/ailia-models/resnet.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:496b6d277f7f05a53a3674fb24d56b5f64e7905af2dc4377c46dee454befc45e
3
+ size 220306739
models/ailia-models/resnet.onnx.prototxt ADDED
@@ -0,0 +1,2023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "model.0.weight"
10
+ input: "model.0.bias"
11
+ output: "65"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 7
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 3
32
+ ints: 3
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 1
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "65"
43
+ input: "model.1.body.0.weight"
44
+ input: "model.1.body.0.bias"
45
+ output: "66"
46
+ name: "Conv_1"
47
+ op_type: "Conv"
48
+ attribute {
49
+ name: "dilations"
50
+ ints: 1
51
+ type: INTS
52
+ }
53
+ attribute {
54
+ name: "group"
55
+ i: 1
56
+ type: INT
57
+ }
58
+ attribute {
59
+ name: "kernel_shape"
60
+ ints: 7
61
+ type: INTS
62
+ }
63
+ attribute {
64
+ name: "pads"
65
+ ints: 3
66
+ ints: 3
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "strides"
71
+ ints: 1
72
+ type: INTS
73
+ }
74
+ }
75
+ node {
76
+ input: "66"
77
+ output: "67"
78
+ name: "Relu_2"
79
+ op_type: "Relu"
80
+ }
81
+ node {
82
+ input: "67"
83
+ input: "model.1.body.2.weight"
84
+ input: "model.1.body.2.bias"
85
+ output: "68"
86
+ name: "Conv_3"
87
+ op_type: "Conv"
88
+ attribute {
89
+ name: "dilations"
90
+ ints: 1
91
+ type: INTS
92
+ }
93
+ attribute {
94
+ name: "group"
95
+ i: 1
96
+ type: INT
97
+ }
98
+ attribute {
99
+ name: "kernel_shape"
100
+ ints: 7
101
+ type: INTS
102
+ }
103
+ attribute {
104
+ name: "pads"
105
+ ints: 3
106
+ ints: 3
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "strides"
111
+ ints: 1
112
+ type: INTS
113
+ }
114
+ }
115
+ node {
116
+ output: "69"
117
+ name: "Constant_4"
118
+ op_type: "Constant"
119
+ attribute {
120
+ name: "value"
121
+ t {
122
+ data_type: 1
123
+ raw_data: "\315\314\314="
124
+ }
125
+ type: TENSOR
126
+ }
127
+ }
128
+ node {
129
+ input: "68"
130
+ input: "69"
131
+ output: "70"
132
+ name: "Mul_5"
133
+ op_type: "Mul"
134
+ }
135
+ node {
136
+ input: "70"
137
+ input: "65"
138
+ output: "71"
139
+ name: "Add_6"
140
+ op_type: "Add"
141
+ }
142
+ node {
143
+ input: "71"
144
+ input: "model.2.body.0.weight"
145
+ input: "model.2.body.0.bias"
146
+ output: "72"
147
+ name: "Conv_7"
148
+ op_type: "Conv"
149
+ attribute {
150
+ name: "dilations"
151
+ ints: 1
152
+ type: INTS
153
+ }
154
+ attribute {
155
+ name: "group"
156
+ i: 1
157
+ type: INT
158
+ }
159
+ attribute {
160
+ name: "kernel_shape"
161
+ ints: 7
162
+ type: INTS
163
+ }
164
+ attribute {
165
+ name: "pads"
166
+ ints: 3
167
+ ints: 3
168
+ type: INTS
169
+ }
170
+ attribute {
171
+ name: "strides"
172
+ ints: 1
173
+ type: INTS
174
+ }
175
+ }
176
+ node {
177
+ input: "72"
178
+ output: "73"
179
+ name: "Relu_8"
180
+ op_type: "Relu"
181
+ }
182
+ node {
183
+ input: "73"
184
+ input: "model.2.body.2.weight"
185
+ input: "model.2.body.2.bias"
186
+ output: "74"
187
+ name: "Conv_9"
188
+ op_type: "Conv"
189
+ attribute {
190
+ name: "dilations"
191
+ ints: 1
192
+ type: INTS
193
+ }
194
+ attribute {
195
+ name: "group"
196
+ i: 1
197
+ type: INT
198
+ }
199
+ attribute {
200
+ name: "kernel_shape"
201
+ ints: 7
202
+ type: INTS
203
+ }
204
+ attribute {
205
+ name: "pads"
206
+ ints: 3
207
+ ints: 3
208
+ type: INTS
209
+ }
210
+ attribute {
211
+ name: "strides"
212
+ ints: 1
213
+ type: INTS
214
+ }
215
+ }
216
+ node {
217
+ output: "75"
218
+ name: "Constant_10"
219
+ op_type: "Constant"
220
+ attribute {
221
+ name: "value"
222
+ t {
223
+ data_type: 1
224
+ raw_data: "\315\314\314="
225
+ }
226
+ type: TENSOR
227
+ }
228
+ }
229
+ node {
230
+ input: "74"
231
+ input: "75"
232
+ output: "76"
233
+ name: "Mul_11"
234
+ op_type: "Mul"
235
+ }
236
+ node {
237
+ input: "76"
238
+ input: "71"
239
+ output: "77"
240
+ name: "Add_12"
241
+ op_type: "Add"
242
+ }
243
+ node {
244
+ input: "77"
245
+ input: "model.3.body.0.weight"
246
+ input: "model.3.body.0.bias"
247
+ output: "78"
248
+ name: "Conv_13"
249
+ op_type: "Conv"
250
+ attribute {
251
+ name: "dilations"
252
+ ints: 1
253
+ type: INTS
254
+ }
255
+ attribute {
256
+ name: "group"
257
+ i: 1
258
+ type: INT
259
+ }
260
+ attribute {
261
+ name: "kernel_shape"
262
+ ints: 7
263
+ type: INTS
264
+ }
265
+ attribute {
266
+ name: "pads"
267
+ ints: 3
268
+ ints: 3
269
+ type: INTS
270
+ }
271
+ attribute {
272
+ name: "strides"
273
+ ints: 1
274
+ type: INTS
275
+ }
276
+ }
277
+ node {
278
+ input: "78"
279
+ output: "79"
280
+ name: "Relu_14"
281
+ op_type: "Relu"
282
+ }
283
+ node {
284
+ input: "79"
285
+ input: "model.3.body.2.weight"
286
+ input: "model.3.body.2.bias"
287
+ output: "80"
288
+ name: "Conv_15"
289
+ op_type: "Conv"
290
+ attribute {
291
+ name: "dilations"
292
+ ints: 1
293
+ type: INTS
294
+ }
295
+ attribute {
296
+ name: "group"
297
+ i: 1
298
+ type: INT
299
+ }
300
+ attribute {
301
+ name: "kernel_shape"
302
+ ints: 7
303
+ type: INTS
304
+ }
305
+ attribute {
306
+ name: "pads"
307
+ ints: 3
308
+ ints: 3
309
+ type: INTS
310
+ }
311
+ attribute {
312
+ name: "strides"
313
+ ints: 1
314
+ type: INTS
315
+ }
316
+ }
317
+ node {
318
+ output: "81"
319
+ name: "Constant_16"
320
+ op_type: "Constant"
321
+ attribute {
322
+ name: "value"
323
+ t {
324
+ data_type: 1
325
+ raw_data: "\315\314\314="
326
+ }
327
+ type: TENSOR
328
+ }
329
+ }
330
+ node {
331
+ input: "80"
332
+ input: "81"
333
+ output: "82"
334
+ name: "Mul_17"
335
+ op_type: "Mul"
336
+ }
337
+ node {
338
+ input: "82"
339
+ input: "77"
340
+ output: "83"
341
+ name: "Add_18"
342
+ op_type: "Add"
343
+ }
344
+ node {
345
+ input: "83"
346
+ input: "model.4.body.0.weight"
347
+ input: "model.4.body.0.bias"
348
+ output: "84"
349
+ name: "Conv_19"
350
+ op_type: "Conv"
351
+ attribute {
352
+ name: "dilations"
353
+ ints: 1
354
+ type: INTS
355
+ }
356
+ attribute {
357
+ name: "group"
358
+ i: 1
359
+ type: INT
360
+ }
361
+ attribute {
362
+ name: "kernel_shape"
363
+ ints: 7
364
+ type: INTS
365
+ }
366
+ attribute {
367
+ name: "pads"
368
+ ints: 3
369
+ ints: 3
370
+ type: INTS
371
+ }
372
+ attribute {
373
+ name: "strides"
374
+ ints: 1
375
+ type: INTS
376
+ }
377
+ }
378
+ node {
379
+ input: "84"
380
+ output: "85"
381
+ name: "Relu_20"
382
+ op_type: "Relu"
383
+ }
384
+ node {
385
+ input: "85"
386
+ input: "model.4.body.2.weight"
387
+ input: "model.4.body.2.bias"
388
+ output: "86"
389
+ name: "Conv_21"
390
+ op_type: "Conv"
391
+ attribute {
392
+ name: "dilations"
393
+ ints: 1
394
+ type: INTS
395
+ }
396
+ attribute {
397
+ name: "group"
398
+ i: 1
399
+ type: INT
400
+ }
401
+ attribute {
402
+ name: "kernel_shape"
403
+ ints: 7
404
+ type: INTS
405
+ }
406
+ attribute {
407
+ name: "pads"
408
+ ints: 3
409
+ ints: 3
410
+ type: INTS
411
+ }
412
+ attribute {
413
+ name: "strides"
414
+ ints: 1
415
+ type: INTS
416
+ }
417
+ }
418
+ node {
419
+ output: "87"
420
+ name: "Constant_22"
421
+ op_type: "Constant"
422
+ attribute {
423
+ name: "value"
424
+ t {
425
+ data_type: 1
426
+ raw_data: "\315\314\314="
427
+ }
428
+ type: TENSOR
429
+ }
430
+ }
431
+ node {
432
+ input: "86"
433
+ input: "87"
434
+ output: "88"
435
+ name: "Mul_23"
436
+ op_type: "Mul"
437
+ }
438
+ node {
439
+ input: "88"
440
+ input: "83"
441
+ output: "89"
442
+ name: "Add_24"
443
+ op_type: "Add"
444
+ }
445
+ node {
446
+ input: "89"
447
+ input: "model.5.body.0.weight"
448
+ input: "model.5.body.0.bias"
449
+ output: "90"
450
+ name: "Conv_25"
451
+ op_type: "Conv"
452
+ attribute {
453
+ name: "dilations"
454
+ ints: 1
455
+ type: INTS
456
+ }
457
+ attribute {
458
+ name: "group"
459
+ i: 1
460
+ type: INT
461
+ }
462
+ attribute {
463
+ name: "kernel_shape"
464
+ ints: 7
465
+ type: INTS
466
+ }
467
+ attribute {
468
+ name: "pads"
469
+ ints: 3
470
+ ints: 3
471
+ type: INTS
472
+ }
473
+ attribute {
474
+ name: "strides"
475
+ ints: 1
476
+ type: INTS
477
+ }
478
+ }
479
+ node {
480
+ input: "90"
481
+ output: "91"
482
+ name: "Relu_26"
483
+ op_type: "Relu"
484
+ }
485
+ node {
486
+ input: "91"
487
+ input: "model.5.body.2.weight"
488
+ input: "model.5.body.2.bias"
489
+ output: "92"
490
+ name: "Conv_27"
491
+ op_type: "Conv"
492
+ attribute {
493
+ name: "dilations"
494
+ ints: 1
495
+ type: INTS
496
+ }
497
+ attribute {
498
+ name: "group"
499
+ i: 1
500
+ type: INT
501
+ }
502
+ attribute {
503
+ name: "kernel_shape"
504
+ ints: 7
505
+ type: INTS
506
+ }
507
+ attribute {
508
+ name: "pads"
509
+ ints: 3
510
+ ints: 3
511
+ type: INTS
512
+ }
513
+ attribute {
514
+ name: "strides"
515
+ ints: 1
516
+ type: INTS
517
+ }
518
+ }
519
+ node {
520
+ output: "93"
521
+ name: "Constant_28"
522
+ op_type: "Constant"
523
+ attribute {
524
+ name: "value"
525
+ t {
526
+ data_type: 1
527
+ raw_data: "\315\314\314="
528
+ }
529
+ type: TENSOR
530
+ }
531
+ }
532
+ node {
533
+ input: "92"
534
+ input: "93"
535
+ output: "94"
536
+ name: "Mul_29"
537
+ op_type: "Mul"
538
+ }
539
+ node {
540
+ input: "94"
541
+ input: "89"
542
+ output: "95"
543
+ name: "Add_30"
544
+ op_type: "Add"
545
+ }
546
+ node {
547
+ input: "95"
548
+ input: "model.6.body.0.weight"
549
+ input: "model.6.body.0.bias"
550
+ output: "96"
551
+ name: "Conv_31"
552
+ op_type: "Conv"
553
+ attribute {
554
+ name: "dilations"
555
+ ints: 1
556
+ type: INTS
557
+ }
558
+ attribute {
559
+ name: "group"
560
+ i: 1
561
+ type: INT
562
+ }
563
+ attribute {
564
+ name: "kernel_shape"
565
+ ints: 7
566
+ type: INTS
567
+ }
568
+ attribute {
569
+ name: "pads"
570
+ ints: 3
571
+ ints: 3
572
+ type: INTS
573
+ }
574
+ attribute {
575
+ name: "strides"
576
+ ints: 1
577
+ type: INTS
578
+ }
579
+ }
580
+ node {
581
+ input: "96"
582
+ output: "97"
583
+ name: "Relu_32"
584
+ op_type: "Relu"
585
+ }
586
+ node {
587
+ input: "97"
588
+ input: "model.6.body.2.weight"
589
+ input: "model.6.body.2.bias"
590
+ output: "98"
591
+ name: "Conv_33"
592
+ op_type: "Conv"
593
+ attribute {
594
+ name: "dilations"
595
+ ints: 1
596
+ type: INTS
597
+ }
598
+ attribute {
599
+ name: "group"
600
+ i: 1
601
+ type: INT
602
+ }
603
+ attribute {
604
+ name: "kernel_shape"
605
+ ints: 7
606
+ type: INTS
607
+ }
608
+ attribute {
609
+ name: "pads"
610
+ ints: 3
611
+ ints: 3
612
+ type: INTS
613
+ }
614
+ attribute {
615
+ name: "strides"
616
+ ints: 1
617
+ type: INTS
618
+ }
619
+ }
620
+ node {
621
+ output: "99"
622
+ name: "Constant_34"
623
+ op_type: "Constant"
624
+ attribute {
625
+ name: "value"
626
+ t {
627
+ data_type: 1
628
+ raw_data: "\315\314\314="
629
+ }
630
+ type: TENSOR
631
+ }
632
+ }
633
+ node {
634
+ input: "98"
635
+ input: "99"
636
+ output: "100"
637
+ name: "Mul_35"
638
+ op_type: "Mul"
639
+ }
640
+ node {
641
+ input: "100"
642
+ input: "95"
643
+ output: "101"
644
+ name: "Add_36"
645
+ op_type: "Add"
646
+ }
647
+ node {
648
+ input: "101"
649
+ input: "model.7.body.0.weight"
650
+ input: "model.7.body.0.bias"
651
+ output: "102"
652
+ name: "Conv_37"
653
+ op_type: "Conv"
654
+ attribute {
655
+ name: "dilations"
656
+ ints: 1
657
+ type: INTS
658
+ }
659
+ attribute {
660
+ name: "group"
661
+ i: 1
662
+ type: INT
663
+ }
664
+ attribute {
665
+ name: "kernel_shape"
666
+ ints: 7
667
+ type: INTS
668
+ }
669
+ attribute {
670
+ name: "pads"
671
+ ints: 3
672
+ ints: 3
673
+ type: INTS
674
+ }
675
+ attribute {
676
+ name: "strides"
677
+ ints: 1
678
+ type: INTS
679
+ }
680
+ }
681
+ node {
682
+ input: "102"
683
+ output: "103"
684
+ name: "Relu_38"
685
+ op_type: "Relu"
686
+ }
687
+ node {
688
+ input: "103"
689
+ input: "model.7.body.2.weight"
690
+ input: "model.7.body.2.bias"
691
+ output: "104"
692
+ name: "Conv_39"
693
+ op_type: "Conv"
694
+ attribute {
695
+ name: "dilations"
696
+ ints: 1
697
+ type: INTS
698
+ }
699
+ attribute {
700
+ name: "group"
701
+ i: 1
702
+ type: INT
703
+ }
704
+ attribute {
705
+ name: "kernel_shape"
706
+ ints: 7
707
+ type: INTS
708
+ }
709
+ attribute {
710
+ name: "pads"
711
+ ints: 3
712
+ ints: 3
713
+ type: INTS
714
+ }
715
+ attribute {
716
+ name: "strides"
717
+ ints: 1
718
+ type: INTS
719
+ }
720
+ }
721
+ node {
722
+ output: "105"
723
+ name: "Constant_40"
724
+ op_type: "Constant"
725
+ attribute {
726
+ name: "value"
727
+ t {
728
+ data_type: 1
729
+ raw_data: "\315\314\314="
730
+ }
731
+ type: TENSOR
732
+ }
733
+ }
734
+ node {
735
+ input: "104"
736
+ input: "105"
737
+ output: "106"
738
+ name: "Mul_41"
739
+ op_type: "Mul"
740
+ }
741
+ node {
742
+ input: "106"
743
+ input: "101"
744
+ output: "107"
745
+ name: "Add_42"
746
+ op_type: "Add"
747
+ }
748
+ node {
749
+ input: "107"
750
+ input: "model.8.body.0.weight"
751
+ input: "model.8.body.0.bias"
752
+ output: "108"
753
+ name: "Conv_43"
754
+ op_type: "Conv"
755
+ attribute {
756
+ name: "dilations"
757
+ ints: 1
758
+ type: INTS
759
+ }
760
+ attribute {
761
+ name: "group"
762
+ i: 1
763
+ type: INT
764
+ }
765
+ attribute {
766
+ name: "kernel_shape"
767
+ ints: 7
768
+ type: INTS
769
+ }
770
+ attribute {
771
+ name: "pads"
772
+ ints: 3
773
+ ints: 3
774
+ type: INTS
775
+ }
776
+ attribute {
777
+ name: "strides"
778
+ ints: 1
779
+ type: INTS
780
+ }
781
+ }
782
+ node {
783
+ input: "108"
784
+ output: "109"
785
+ name: "Relu_44"
786
+ op_type: "Relu"
787
+ }
788
+ node {
789
+ input: "109"
790
+ input: "model.8.body.2.weight"
791
+ input: "model.8.body.2.bias"
792
+ output: "110"
793
+ name: "Conv_45"
794
+ op_type: "Conv"
795
+ attribute {
796
+ name: "dilations"
797
+ ints: 1
798
+ type: INTS
799
+ }
800
+ attribute {
801
+ name: "group"
802
+ i: 1
803
+ type: INT
804
+ }
805
+ attribute {
806
+ name: "kernel_shape"
807
+ ints: 7
808
+ type: INTS
809
+ }
810
+ attribute {
811
+ name: "pads"
812
+ ints: 3
813
+ ints: 3
814
+ type: INTS
815
+ }
816
+ attribute {
817
+ name: "strides"
818
+ ints: 1
819
+ type: INTS
820
+ }
821
+ }
822
+ node {
823
+ output: "111"
824
+ name: "Constant_46"
825
+ op_type: "Constant"
826
+ attribute {
827
+ name: "value"
828
+ t {
829
+ data_type: 1
830
+ raw_data: "\315\314\314="
831
+ }
832
+ type: TENSOR
833
+ }
834
+ }
835
+ node {
836
+ input: "110"
837
+ input: "111"
838
+ output: "112"
839
+ name: "Mul_47"
840
+ op_type: "Mul"
841
+ }
842
+ node {
843
+ input: "112"
844
+ input: "107"
845
+ output: "113"
846
+ name: "Add_48"
847
+ op_type: "Add"
848
+ }
849
+ node {
850
+ input: "113"
851
+ input: "model.9.body.0.weight"
852
+ input: "model.9.body.0.bias"
853
+ output: "114"
854
+ name: "Conv_49"
855
+ op_type: "Conv"
856
+ attribute {
857
+ name: "dilations"
858
+ ints: 1
859
+ type: INTS
860
+ }
861
+ attribute {
862
+ name: "group"
863
+ i: 1
864
+ type: INT
865
+ }
866
+ attribute {
867
+ name: "kernel_shape"
868
+ ints: 7
869
+ type: INTS
870
+ }
871
+ attribute {
872
+ name: "pads"
873
+ ints: 3
874
+ ints: 3
875
+ type: INTS
876
+ }
877
+ attribute {
878
+ name: "strides"
879
+ ints: 1
880
+ type: INTS
881
+ }
882
+ }
883
+ node {
884
+ input: "114"
885
+ output: "115"
886
+ name: "Relu_50"
887
+ op_type: "Relu"
888
+ }
889
+ node {
890
+ input: "115"
891
+ input: "model.9.body.2.weight"
892
+ input: "model.9.body.2.bias"
893
+ output: "116"
894
+ name: "Conv_51"
895
+ op_type: "Conv"
896
+ attribute {
897
+ name: "dilations"
898
+ ints: 1
899
+ type: INTS
900
+ }
901
+ attribute {
902
+ name: "group"
903
+ i: 1
904
+ type: INT
905
+ }
906
+ attribute {
907
+ name: "kernel_shape"
908
+ ints: 7
909
+ type: INTS
910
+ }
911
+ attribute {
912
+ name: "pads"
913
+ ints: 3
914
+ ints: 3
915
+ type: INTS
916
+ }
917
+ attribute {
918
+ name: "strides"
919
+ ints: 1
920
+ type: INTS
921
+ }
922
+ }
923
+ node {
924
+ output: "117"
925
+ name: "Constant_52"
926
+ op_type: "Constant"
927
+ attribute {
928
+ name: "value"
929
+ t {
930
+ data_type: 1
931
+ raw_data: "\315\314\314="
932
+ }
933
+ type: TENSOR
934
+ }
935
+ }
936
+ node {
937
+ input: "116"
938
+ input: "117"
939
+ output: "118"
940
+ name: "Mul_53"
941
+ op_type: "Mul"
942
+ }
943
+ node {
944
+ input: "118"
945
+ input: "113"
946
+ output: "119"
947
+ name: "Add_54"
948
+ op_type: "Add"
949
+ }
950
+ node {
951
+ input: "119"
952
+ input: "model.10.body.0.weight"
953
+ input: "model.10.body.0.bias"
954
+ output: "120"
955
+ name: "Conv_55"
956
+ op_type: "Conv"
957
+ attribute {
958
+ name: "dilations"
959
+ ints: 1
960
+ type: INTS
961
+ }
962
+ attribute {
963
+ name: "group"
964
+ i: 1
965
+ type: INT
966
+ }
967
+ attribute {
968
+ name: "kernel_shape"
969
+ ints: 7
970
+ type: INTS
971
+ }
972
+ attribute {
973
+ name: "pads"
974
+ ints: 3
975
+ ints: 3
976
+ type: INTS
977
+ }
978
+ attribute {
979
+ name: "strides"
980
+ ints: 1
981
+ type: INTS
982
+ }
983
+ }
984
+ node {
985
+ input: "120"
986
+ output: "121"
987
+ name: "Relu_56"
988
+ op_type: "Relu"
989
+ }
990
+ node {
991
+ input: "121"
992
+ input: "model.10.body.2.weight"
993
+ input: "model.10.body.2.bias"
994
+ output: "122"
995
+ name: "Conv_57"
996
+ op_type: "Conv"
997
+ attribute {
998
+ name: "dilations"
999
+ ints: 1
1000
+ type: INTS
1001
+ }
1002
+ attribute {
1003
+ name: "group"
1004
+ i: 1
1005
+ type: INT
1006
+ }
1007
+ attribute {
1008
+ name: "kernel_shape"
1009
+ ints: 7
1010
+ type: INTS
1011
+ }
1012
+ attribute {
1013
+ name: "pads"
1014
+ ints: 3
1015
+ ints: 3
1016
+ type: INTS
1017
+ }
1018
+ attribute {
1019
+ name: "strides"
1020
+ ints: 1
1021
+ type: INTS
1022
+ }
1023
+ }
1024
+ node {
1025
+ output: "123"
1026
+ name: "Constant_58"
1027
+ op_type: "Constant"
1028
+ attribute {
1029
+ name: "value"
1030
+ t {
1031
+ data_type: 1
1032
+ raw_data: "\315\314\314="
1033
+ }
1034
+ type: TENSOR
1035
+ }
1036
+ }
1037
+ node {
1038
+ input: "122"
1039
+ input: "123"
1040
+ output: "124"
1041
+ name: "Mul_59"
1042
+ op_type: "Mul"
1043
+ }
1044
+ node {
1045
+ input: "124"
1046
+ input: "119"
1047
+ output: "125"
1048
+ name: "Add_60"
1049
+ op_type: "Add"
1050
+ }
1051
+ node {
1052
+ input: "125"
1053
+ input: "model.11.body.0.weight"
1054
+ input: "model.11.body.0.bias"
1055
+ output: "126"
1056
+ name: "Conv_61"
1057
+ op_type: "Conv"
1058
+ attribute {
1059
+ name: "dilations"
1060
+ ints: 1
1061
+ type: INTS
1062
+ }
1063
+ attribute {
1064
+ name: "group"
1065
+ i: 1
1066
+ type: INT
1067
+ }
1068
+ attribute {
1069
+ name: "kernel_shape"
1070
+ ints: 7
1071
+ type: INTS
1072
+ }
1073
+ attribute {
1074
+ name: "pads"
1075
+ ints: 3
1076
+ ints: 3
1077
+ type: INTS
1078
+ }
1079
+ attribute {
1080
+ name: "strides"
1081
+ ints: 1
1082
+ type: INTS
1083
+ }
1084
+ }
1085
+ node {
1086
+ input: "126"
1087
+ output: "127"
1088
+ name: "Relu_62"
1089
+ op_type: "Relu"
1090
+ }
1091
+ node {
1092
+ input: "127"
1093
+ input: "model.11.body.2.weight"
1094
+ input: "model.11.body.2.bias"
1095
+ output: "128"
1096
+ name: "Conv_63"
1097
+ op_type: "Conv"
1098
+ attribute {
1099
+ name: "dilations"
1100
+ ints: 1
1101
+ type: INTS
1102
+ }
1103
+ attribute {
1104
+ name: "group"
1105
+ i: 1
1106
+ type: INT
1107
+ }
1108
+ attribute {
1109
+ name: "kernel_shape"
1110
+ ints: 7
1111
+ type: INTS
1112
+ }
1113
+ attribute {
1114
+ name: "pads"
1115
+ ints: 3
1116
+ ints: 3
1117
+ type: INTS
1118
+ }
1119
+ attribute {
1120
+ name: "strides"
1121
+ ints: 1
1122
+ type: INTS
1123
+ }
1124
+ }
1125
+ node {
1126
+ output: "129"
1127
+ name: "Constant_64"
1128
+ op_type: "Constant"
1129
+ attribute {
1130
+ name: "value"
1131
+ t {
1132
+ data_type: 1
1133
+ raw_data: "\315\314\314="
1134
+ }
1135
+ type: TENSOR
1136
+ }
1137
+ }
1138
+ node {
1139
+ input: "128"
1140
+ input: "129"
1141
+ output: "130"
1142
+ name: "Mul_65"
1143
+ op_type: "Mul"
1144
+ }
1145
+ node {
1146
+ input: "130"
1147
+ input: "125"
1148
+ output: "131"
1149
+ name: "Add_66"
1150
+ op_type: "Add"
1151
+ }
1152
+ node {
1153
+ input: "131"
1154
+ input: "model.12.body.0.weight"
1155
+ input: "model.12.body.0.bias"
1156
+ output: "132"
1157
+ name: "Conv_67"
1158
+ op_type: "Conv"
1159
+ attribute {
1160
+ name: "dilations"
1161
+ ints: 1
1162
+ type: INTS
1163
+ }
1164
+ attribute {
1165
+ name: "group"
1166
+ i: 1
1167
+ type: INT
1168
+ }
1169
+ attribute {
1170
+ name: "kernel_shape"
1171
+ ints: 7
1172
+ type: INTS
1173
+ }
1174
+ attribute {
1175
+ name: "pads"
1176
+ ints: 3
1177
+ ints: 3
1178
+ type: INTS
1179
+ }
1180
+ attribute {
1181
+ name: "strides"
1182
+ ints: 1
1183
+ type: INTS
1184
+ }
1185
+ }
1186
+ node {
1187
+ input: "132"
1188
+ output: "133"
1189
+ name: "Relu_68"
1190
+ op_type: "Relu"
1191
+ }
1192
+ node {
1193
+ input: "133"
1194
+ input: "model.12.body.2.weight"
1195
+ input: "model.12.body.2.bias"
1196
+ output: "134"
1197
+ name: "Conv_69"
1198
+ op_type: "Conv"
1199
+ attribute {
1200
+ name: "dilations"
1201
+ ints: 1
1202
+ type: INTS
1203
+ }
1204
+ attribute {
1205
+ name: "group"
1206
+ i: 1
1207
+ type: INT
1208
+ }
1209
+ attribute {
1210
+ name: "kernel_shape"
1211
+ ints: 7
1212
+ type: INTS
1213
+ }
1214
+ attribute {
1215
+ name: "pads"
1216
+ ints: 3
1217
+ ints: 3
1218
+ type: INTS
1219
+ }
1220
+ attribute {
1221
+ name: "strides"
1222
+ ints: 1
1223
+ type: INTS
1224
+ }
1225
+ }
1226
+ node {
1227
+ output: "135"
1228
+ name: "Constant_70"
1229
+ op_type: "Constant"
1230
+ attribute {
1231
+ name: "value"
1232
+ t {
1233
+ data_type: 1
1234
+ raw_data: "\315\314\314="
1235
+ }
1236
+ type: TENSOR
1237
+ }
1238
+ }
1239
+ node {
1240
+ input: "134"
1241
+ input: "135"
1242
+ output: "136"
1243
+ name: "Mul_71"
1244
+ op_type: "Mul"
1245
+ }
1246
+ node {
1247
+ input: "136"
1248
+ input: "131"
1249
+ output: "137"
1250
+ name: "Add_72"
1251
+ op_type: "Add"
1252
+ }
1253
+ node {
1254
+ input: "137"
1255
+ input: "model.13.body.0.weight"
1256
+ input: "model.13.body.0.bias"
1257
+ output: "138"
1258
+ name: "Conv_73"
1259
+ op_type: "Conv"
1260
+ attribute {
1261
+ name: "dilations"
1262
+ ints: 1
1263
+ type: INTS
1264
+ }
1265
+ attribute {
1266
+ name: "group"
1267
+ i: 1
1268
+ type: INT
1269
+ }
1270
+ attribute {
1271
+ name: "kernel_shape"
1272
+ ints: 7
1273
+ type: INTS
1274
+ }
1275
+ attribute {
1276
+ name: "pads"
1277
+ ints: 3
1278
+ ints: 3
1279
+ type: INTS
1280
+ }
1281
+ attribute {
1282
+ name: "strides"
1283
+ ints: 1
1284
+ type: INTS
1285
+ }
1286
+ }
1287
+ node {
1288
+ input: "138"
1289
+ output: "139"
1290
+ name: "Relu_74"
1291
+ op_type: "Relu"
1292
+ }
1293
+ node {
1294
+ input: "139"
1295
+ input: "model.13.body.2.weight"
1296
+ input: "model.13.body.2.bias"
1297
+ output: "140"
1298
+ name: "Conv_75"
1299
+ op_type: "Conv"
1300
+ attribute {
1301
+ name: "dilations"
1302
+ ints: 1
1303
+ type: INTS
1304
+ }
1305
+ attribute {
1306
+ name: "group"
1307
+ i: 1
1308
+ type: INT
1309
+ }
1310
+ attribute {
1311
+ name: "kernel_shape"
1312
+ ints: 7
1313
+ type: INTS
1314
+ }
1315
+ attribute {
1316
+ name: "pads"
1317
+ ints: 3
1318
+ ints: 3
1319
+ type: INTS
1320
+ }
1321
+ attribute {
1322
+ name: "strides"
1323
+ ints: 1
1324
+ type: INTS
1325
+ }
1326
+ }
1327
+ node {
1328
+ output: "141"
1329
+ name: "Constant_76"
1330
+ op_type: "Constant"
1331
+ attribute {
1332
+ name: "value"
1333
+ t {
1334
+ data_type: 1
1335
+ raw_data: "\315\314\314="
1336
+ }
1337
+ type: TENSOR
1338
+ }
1339
+ }
1340
+ node {
1341
+ input: "140"
1342
+ input: "141"
1343
+ output: "142"
1344
+ name: "Mul_77"
1345
+ op_type: "Mul"
1346
+ }
1347
+ node {
1348
+ input: "142"
1349
+ input: "137"
1350
+ output: "143"
1351
+ name: "Add_78"
1352
+ op_type: "Add"
1353
+ }
1354
+ node {
1355
+ input: "143"
1356
+ input: "model.14.body.0.weight"
1357
+ input: "model.14.body.0.bias"
1358
+ output: "144"
1359
+ name: "Conv_79"
1360
+ op_type: "Conv"
1361
+ attribute {
1362
+ name: "dilations"
1363
+ ints: 1
1364
+ type: INTS
1365
+ }
1366
+ attribute {
1367
+ name: "group"
1368
+ i: 1
1369
+ type: INT
1370
+ }
1371
+ attribute {
1372
+ name: "kernel_shape"
1373
+ ints: 7
1374
+ type: INTS
1375
+ }
1376
+ attribute {
1377
+ name: "pads"
1378
+ ints: 3
1379
+ ints: 3
1380
+ type: INTS
1381
+ }
1382
+ attribute {
1383
+ name: "strides"
1384
+ ints: 1
1385
+ type: INTS
1386
+ }
1387
+ }
1388
+ node {
1389
+ input: "144"
1390
+ output: "145"
1391
+ name: "Relu_80"
1392
+ op_type: "Relu"
1393
+ }
1394
+ node {
1395
+ input: "145"
1396
+ input: "model.14.body.2.weight"
1397
+ input: "model.14.body.2.bias"
1398
+ output: "146"
1399
+ name: "Conv_81"
1400
+ op_type: "Conv"
1401
+ attribute {
1402
+ name: "dilations"
1403
+ ints: 1
1404
+ type: INTS
1405
+ }
1406
+ attribute {
1407
+ name: "group"
1408
+ i: 1
1409
+ type: INT
1410
+ }
1411
+ attribute {
1412
+ name: "kernel_shape"
1413
+ ints: 7
1414
+ type: INTS
1415
+ }
1416
+ attribute {
1417
+ name: "pads"
1418
+ ints: 3
1419
+ ints: 3
1420
+ type: INTS
1421
+ }
1422
+ attribute {
1423
+ name: "strides"
1424
+ ints: 1
1425
+ type: INTS
1426
+ }
1427
+ }
1428
+ node {
1429
+ output: "147"
1430
+ name: "Constant_82"
1431
+ op_type: "Constant"
1432
+ attribute {
1433
+ name: "value"
1434
+ t {
1435
+ data_type: 1
1436
+ raw_data: "\315\314\314="
1437
+ }
1438
+ type: TENSOR
1439
+ }
1440
+ }
1441
+ node {
1442
+ input: "146"
1443
+ input: "147"
1444
+ output: "148"
1445
+ name: "Mul_83"
1446
+ op_type: "Mul"
1447
+ }
1448
+ node {
1449
+ input: "148"
1450
+ input: "143"
1451
+ output: "149"
1452
+ name: "Add_84"
1453
+ op_type: "Add"
1454
+ }
1455
+ node {
1456
+ input: "149"
1457
+ input: "model.15.body.0.weight"
1458
+ input: "model.15.body.0.bias"
1459
+ output: "150"
1460
+ name: "Conv_85"
1461
+ op_type: "Conv"
1462
+ attribute {
1463
+ name: "dilations"
1464
+ ints: 1
1465
+ type: INTS
1466
+ }
1467
+ attribute {
1468
+ name: "group"
1469
+ i: 1
1470
+ type: INT
1471
+ }
1472
+ attribute {
1473
+ name: "kernel_shape"
1474
+ ints: 7
1475
+ type: INTS
1476
+ }
1477
+ attribute {
1478
+ name: "pads"
1479
+ ints: 3
1480
+ ints: 3
1481
+ type: INTS
1482
+ }
1483
+ attribute {
1484
+ name: "strides"
1485
+ ints: 1
1486
+ type: INTS
1487
+ }
1488
+ }
1489
+ node {
1490
+ input: "150"
1491
+ output: "151"
1492
+ name: "Relu_86"
1493
+ op_type: "Relu"
1494
+ }
1495
+ node {
1496
+ input: "151"
1497
+ input: "model.15.body.2.weight"
1498
+ input: "model.15.body.2.bias"
1499
+ output: "152"
1500
+ name: "Conv_87"
1501
+ op_type: "Conv"
1502
+ attribute {
1503
+ name: "dilations"
1504
+ ints: 1
1505
+ type: INTS
1506
+ }
1507
+ attribute {
1508
+ name: "group"
1509
+ i: 1
1510
+ type: INT
1511
+ }
1512
+ attribute {
1513
+ name: "kernel_shape"
1514
+ ints: 7
1515
+ type: INTS
1516
+ }
1517
+ attribute {
1518
+ name: "pads"
1519
+ ints: 3
1520
+ ints: 3
1521
+ type: INTS
1522
+ }
1523
+ attribute {
1524
+ name: "strides"
1525
+ ints: 1
1526
+ type: INTS
1527
+ }
1528
+ }
1529
+ node {
1530
+ output: "153"
1531
+ name: "Constant_88"
1532
+ op_type: "Constant"
1533
+ attribute {
1534
+ name: "value"
1535
+ t {
1536
+ data_type: 1
1537
+ raw_data: "\315\314\314="
1538
+ }
1539
+ type: TENSOR
1540
+ }
1541
+ }
1542
+ node {
1543
+ input: "152"
1544
+ input: "153"
1545
+ output: "154"
1546
+ name: "Mul_89"
1547
+ op_type: "Mul"
1548
+ }
1549
+ node {
1550
+ input: "154"
1551
+ input: "149"
1552
+ output: "155"
1553
+ name: "Add_90"
1554
+ op_type: "Add"
1555
+ }
1556
+ node {
1557
+ input: "155"
1558
+ input: "model.16.weight"
1559
+ input: "model.16.bias"
1560
+ output: "156"
1561
+ name: "Conv_91"
1562
+ op_type: "Conv"
1563
+ attribute {
1564
+ name: "dilations"
1565
+ ints: 1
1566
+ type: INTS
1567
+ }
1568
+ attribute {
1569
+ name: "group"
1570
+ i: 1
1571
+ type: INT
1572
+ }
1573
+ attribute {
1574
+ name: "kernel_shape"
1575
+ ints: 1
1576
+ type: INTS
1577
+ }
1578
+ attribute {
1579
+ name: "pads"
1580
+ ints: 0
1581
+ ints: 0
1582
+ type: INTS
1583
+ }
1584
+ attribute {
1585
+ name: "strides"
1586
+ ints: 1
1587
+ type: INTS
1588
+ }
1589
+ }
1590
+ node {
1591
+ input: "156"
1592
+ input: "x"
1593
+ output: "y"
1594
+ name: "Add_92"
1595
+ op_type: "Add"
1596
+ }
1597
+ initializer {
1598
+ dims: 512
1599
+ data_type: 1
1600
+ name: "model.0.bias"
1601
+ }
1602
+ initializer {
1603
+ dims: 512
1604
+ dims: 2
1605
+ dims: 7
1606
+ data_type: 1
1607
+ name: "model.0.weight"
1608
+ }
1609
+ initializer {
1610
+ dims: 512
1611
+ data_type: 1
1612
+ name: "model.1.body.0.bias"
1613
+ }
1614
+ initializer {
1615
+ dims: 512
1616
+ dims: 512
1617
+ dims: 7
1618
+ data_type: 1
1619
+ name: "model.1.body.0.weight"
1620
+ }
1621
+ initializer {
1622
+ dims: 512
1623
+ data_type: 1
1624
+ name: "model.1.body.2.bias"
1625
+ }
1626
+ initializer {
1627
+ dims: 512
1628
+ dims: 512
1629
+ dims: 7
1630
+ data_type: 1
1631
+ name: "model.1.body.2.weight"
1632
+ }
1633
+ initializer {
1634
+ dims: 512
1635
+ data_type: 1
1636
+ name: "model.10.body.0.bias"
1637
+ }
1638
+ initializer {
1639
+ dims: 512
1640
+ dims: 512
1641
+ dims: 7
1642
+ data_type: 1
1643
+ name: "model.10.body.0.weight"
1644
+ }
1645
+ initializer {
1646
+ dims: 512
1647
+ data_type: 1
1648
+ name: "model.10.body.2.bias"
1649
+ }
1650
+ initializer {
1651
+ dims: 512
1652
+ dims: 512
1653
+ dims: 7
1654
+ data_type: 1
1655
+ name: "model.10.body.2.weight"
1656
+ }
1657
+ initializer {
1658
+ dims: 512
1659
+ data_type: 1
1660
+ name: "model.11.body.0.bias"
1661
+ }
1662
+ initializer {
1663
+ dims: 512
1664
+ dims: 512
1665
+ dims: 7
1666
+ data_type: 1
1667
+ name: "model.11.body.0.weight"
1668
+ }
1669
+ initializer {
1670
+ dims: 512
1671
+ data_type: 1
1672
+ name: "model.11.body.2.bias"
1673
+ }
1674
+ initializer {
1675
+ dims: 512
1676
+ dims: 512
1677
+ dims: 7
1678
+ data_type: 1
1679
+ name: "model.11.body.2.weight"
1680
+ }
1681
+ initializer {
1682
+ dims: 512
1683
+ data_type: 1
1684
+ name: "model.12.body.0.bias"
1685
+ }
1686
+ initializer {
1687
+ dims: 512
1688
+ dims: 512
1689
+ dims: 7
1690
+ data_type: 1
1691
+ name: "model.12.body.0.weight"
1692
+ }
1693
+ initializer {
1694
+ dims: 512
1695
+ data_type: 1
1696
+ name: "model.12.body.2.bias"
1697
+ }
1698
+ initializer {
1699
+ dims: 512
1700
+ dims: 512
1701
+ dims: 7
1702
+ data_type: 1
1703
+ name: "model.12.body.2.weight"
1704
+ }
1705
+ initializer {
1706
+ dims: 512
1707
+ data_type: 1
1708
+ name: "model.13.body.0.bias"
1709
+ }
1710
+ initializer {
1711
+ dims: 512
1712
+ dims: 512
1713
+ dims: 7
1714
+ data_type: 1
1715
+ name: "model.13.body.0.weight"
1716
+ }
1717
+ initializer {
1718
+ dims: 512
1719
+ data_type: 1
1720
+ name: "model.13.body.2.bias"
1721
+ }
1722
+ initializer {
1723
+ dims: 512
1724
+ dims: 512
1725
+ dims: 7
1726
+ data_type: 1
1727
+ name: "model.13.body.2.weight"
1728
+ }
1729
+ initializer {
1730
+ dims: 512
1731
+ data_type: 1
1732
+ name: "model.14.body.0.bias"
1733
+ }
1734
+ initializer {
1735
+ dims: 512
1736
+ dims: 512
1737
+ dims: 7
1738
+ data_type: 1
1739
+ name: "model.14.body.0.weight"
1740
+ }
1741
+ initializer {
1742
+ dims: 512
1743
+ data_type: 1
1744
+ name: "model.14.body.2.bias"
1745
+ }
1746
+ initializer {
1747
+ dims: 512
1748
+ dims: 512
1749
+ dims: 7
1750
+ data_type: 1
1751
+ name: "model.14.body.2.weight"
1752
+ }
1753
+ initializer {
1754
+ dims: 512
1755
+ data_type: 1
1756
+ name: "model.15.body.0.bias"
1757
+ }
1758
+ initializer {
1759
+ dims: 512
1760
+ dims: 512
1761
+ dims: 7
1762
+ data_type: 1
1763
+ name: "model.15.body.0.weight"
1764
+ }
1765
+ initializer {
1766
+ dims: 512
1767
+ data_type: 1
1768
+ name: "model.15.body.2.bias"
1769
+ }
1770
+ initializer {
1771
+ dims: 512
1772
+ dims: 512
1773
+ dims: 7
1774
+ data_type: 1
1775
+ name: "model.15.body.2.weight"
1776
+ }
1777
+ initializer {
1778
+ dims: 2
1779
+ data_type: 1
1780
+ name: "model.16.bias"
1781
+ raw_data: "\263\377\363<?\323\016="
1782
+ }
1783
+ initializer {
1784
+ dims: 2
1785
+ dims: 512
1786
+ dims: 1
1787
+ data_type: 1
1788
+ name: "model.16.weight"
1789
+ }
1790
+ initializer {
1791
+ dims: 512
1792
+ data_type: 1
1793
+ name: "model.2.body.0.bias"
1794
+ }
1795
+ initializer {
1796
+ dims: 512
1797
+ dims: 512
1798
+ dims: 7
1799
+ data_type: 1
1800
+ name: "model.2.body.0.weight"
1801
+ }
1802
+ initializer {
1803
+ dims: 512
1804
+ data_type: 1
1805
+ name: "model.2.body.2.bias"
1806
+ }
1807
+ initializer {
1808
+ dims: 512
1809
+ dims: 512
1810
+ dims: 7
1811
+ data_type: 1
1812
+ name: "model.2.body.2.weight"
1813
+ }
1814
+ initializer {
1815
+ dims: 512
1816
+ data_type: 1
1817
+ name: "model.3.body.0.bias"
1818
+ }
1819
+ initializer {
1820
+ dims: 512
1821
+ dims: 512
1822
+ dims: 7
1823
+ data_type: 1
1824
+ name: "model.3.body.0.weight"
1825
+ }
1826
+ initializer {
1827
+ dims: 512
1828
+ data_type: 1
1829
+ name: "model.3.body.2.bias"
1830
+ }
1831
+ initializer {
1832
+ dims: 512
1833
+ dims: 512
1834
+ dims: 7
1835
+ data_type: 1
1836
+ name: "model.3.body.2.weight"
1837
+ }
1838
+ initializer {
1839
+ dims: 512
1840
+ data_type: 1
1841
+ name: "model.4.body.0.bias"
1842
+ }
1843
+ initializer {
1844
+ dims: 512
1845
+ dims: 512
1846
+ dims: 7
1847
+ data_type: 1
1848
+ name: "model.4.body.0.weight"
1849
+ }
1850
+ initializer {
1851
+ dims: 512
1852
+ data_type: 1
1853
+ name: "model.4.body.2.bias"
1854
+ }
1855
+ initializer {
1856
+ dims: 512
1857
+ dims: 512
1858
+ dims: 7
1859
+ data_type: 1
1860
+ name: "model.4.body.2.weight"
1861
+ }
1862
+ initializer {
1863
+ dims: 512
1864
+ data_type: 1
1865
+ name: "model.5.body.0.bias"
1866
+ }
1867
+ initializer {
1868
+ dims: 512
1869
+ dims: 512
1870
+ dims: 7
1871
+ data_type: 1
1872
+ name: "model.5.body.0.weight"
1873
+ }
1874
+ initializer {
1875
+ dims: 512
1876
+ data_type: 1
1877
+ name: "model.5.body.2.bias"
1878
+ }
1879
+ initializer {
1880
+ dims: 512
1881
+ dims: 512
1882
+ dims: 7
1883
+ data_type: 1
1884
+ name: "model.5.body.2.weight"
1885
+ }
1886
+ initializer {
1887
+ dims: 512
1888
+ data_type: 1
1889
+ name: "model.6.body.0.bias"
1890
+ }
1891
+ initializer {
1892
+ dims: 512
1893
+ dims: 512
1894
+ dims: 7
1895
+ data_type: 1
1896
+ name: "model.6.body.0.weight"
1897
+ }
1898
+ initializer {
1899
+ dims: 512
1900
+ data_type: 1
1901
+ name: "model.6.body.2.bias"
1902
+ }
1903
+ initializer {
1904
+ dims: 512
1905
+ dims: 512
1906
+ dims: 7
1907
+ data_type: 1
1908
+ name: "model.6.body.2.weight"
1909
+ }
1910
+ initializer {
1911
+ dims: 512
1912
+ data_type: 1
1913
+ name: "model.7.body.0.bias"
1914
+ }
1915
+ initializer {
1916
+ dims: 512
1917
+ dims: 512
1918
+ dims: 7
1919
+ data_type: 1
1920
+ name: "model.7.body.0.weight"
1921
+ }
1922
+ initializer {
1923
+ dims: 512
1924
+ data_type: 1
1925
+ name: "model.7.body.2.bias"
1926
+ }
1927
+ initializer {
1928
+ dims: 512
1929
+ dims: 512
1930
+ dims: 7
1931
+ data_type: 1
1932
+ name: "model.7.body.2.weight"
1933
+ }
1934
+ initializer {
1935
+ dims: 512
1936
+ data_type: 1
1937
+ name: "model.8.body.0.bias"
1938
+ }
1939
+ initializer {
1940
+ dims: 512
1941
+ dims: 512
1942
+ dims: 7
1943
+ data_type: 1
1944
+ name: "model.8.body.0.weight"
1945
+ }
1946
+ initializer {
1947
+ dims: 512
1948
+ data_type: 1
1949
+ name: "model.8.body.2.bias"
1950
+ }
1951
+ initializer {
1952
+ dims: 512
1953
+ dims: 512
1954
+ dims: 7
1955
+ data_type: 1
1956
+ name: "model.8.body.2.weight"
1957
+ }
1958
+ initializer {
1959
+ dims: 512
1960
+ data_type: 1
1961
+ name: "model.9.body.0.bias"
1962
+ }
1963
+ initializer {
1964
+ dims: 512
1965
+ dims: 512
1966
+ dims: 7
1967
+ data_type: 1
1968
+ name: "model.9.body.0.weight"
1969
+ }
1970
+ initializer {
1971
+ dims: 512
1972
+ data_type: 1
1973
+ name: "model.9.body.2.bias"
1974
+ }
1975
+ initializer {
1976
+ dims: 512
1977
+ dims: 512
1978
+ dims: 7
1979
+ data_type: 1
1980
+ name: "model.9.body.2.weight"
1981
+ }
1982
+ input {
1983
+ name: "x"
1984
+ type {
1985
+ tensor_type {
1986
+ elem_type: 1
1987
+ shape {
1988
+ dim {
1989
+ dim_param: "batch_size"
1990
+ }
1991
+ dim {
1992
+ dim_value: 2
1993
+ }
1994
+ dim {
1995
+ dim_value: 8192
1996
+ }
1997
+ }
1998
+ }
1999
+ }
2000
+ }
2001
+ output {
2002
+ name: "y"
2003
+ type {
2004
+ tensor_type {
2005
+ elem_type: 1
2006
+ shape {
2007
+ dim {
2008
+ dim_param: "batch_size"
2009
+ }
2010
+ dim {
2011
+ dim_value: 2
2012
+ }
2013
+ dim {
2014
+ dim_value: 8192
2015
+ }
2016
+ }
2017
+ }
2018
+ }
2019
+ }
2020
+ }
2021
+ opset_import {
2022
+ version: 11
2023
+ }
models/ailia-models/resnetbn.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1857ce859375c7c283209706c6aac1381bf93ee441c054b2c22f56434d6fdd8
3
+ size 220304697
models/ailia-models/resnetbn.onnx.prototxt ADDED
@@ -0,0 +1,2023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "model.0.weight"
10
+ input: "model.0.bias"
11
+ output: "185"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 7
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 3
32
+ ints: 3
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 1
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "185"
43
+ input: "309"
44
+ input: "310"
45
+ output: "308"
46
+ name: "Conv_1"
47
+ op_type: "Conv"
48
+ attribute {
49
+ name: "dilations"
50
+ ints: 1
51
+ type: INTS
52
+ }
53
+ attribute {
54
+ name: "group"
55
+ i: 1
56
+ type: INT
57
+ }
58
+ attribute {
59
+ name: "kernel_shape"
60
+ ints: 7
61
+ type: INTS
62
+ }
63
+ attribute {
64
+ name: "pads"
65
+ ints: 3
66
+ ints: 3
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "strides"
71
+ ints: 1
72
+ type: INTS
73
+ }
74
+ }
75
+ node {
76
+ input: "308"
77
+ output: "188"
78
+ name: "Relu_2"
79
+ op_type: "Relu"
80
+ }
81
+ node {
82
+ input: "188"
83
+ input: "312"
84
+ input: "313"
85
+ output: "311"
86
+ name: "Conv_3"
87
+ op_type: "Conv"
88
+ attribute {
89
+ name: "dilations"
90
+ ints: 1
91
+ type: INTS
92
+ }
93
+ attribute {
94
+ name: "group"
95
+ i: 1
96
+ type: INT
97
+ }
98
+ attribute {
99
+ name: "kernel_shape"
100
+ ints: 7
101
+ type: INTS
102
+ }
103
+ attribute {
104
+ name: "pads"
105
+ ints: 3
106
+ ints: 3
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "strides"
111
+ ints: 1
112
+ type: INTS
113
+ }
114
+ }
115
+ node {
116
+ output: "191"
117
+ name: "Constant_4"
118
+ op_type: "Constant"
119
+ attribute {
120
+ name: "value"
121
+ t {
122
+ data_type: 1
123
+ raw_data: "\315\314\314="
124
+ }
125
+ type: TENSOR
126
+ }
127
+ }
128
+ node {
129
+ input: "311"
130
+ input: "191"
131
+ output: "192"
132
+ name: "Mul_5"
133
+ op_type: "Mul"
134
+ }
135
+ node {
136
+ input: "192"
137
+ input: "185"
138
+ output: "193"
139
+ name: "Add_6"
140
+ op_type: "Add"
141
+ }
142
+ node {
143
+ input: "193"
144
+ input: "315"
145
+ input: "316"
146
+ output: "314"
147
+ name: "Conv_7"
148
+ op_type: "Conv"
149
+ attribute {
150
+ name: "dilations"
151
+ ints: 1
152
+ type: INTS
153
+ }
154
+ attribute {
155
+ name: "group"
156
+ i: 1
157
+ type: INT
158
+ }
159
+ attribute {
160
+ name: "kernel_shape"
161
+ ints: 7
162
+ type: INTS
163
+ }
164
+ attribute {
165
+ name: "pads"
166
+ ints: 3
167
+ ints: 3
168
+ type: INTS
169
+ }
170
+ attribute {
171
+ name: "strides"
172
+ ints: 1
173
+ type: INTS
174
+ }
175
+ }
176
+ node {
177
+ input: "314"
178
+ output: "196"
179
+ name: "Relu_8"
180
+ op_type: "Relu"
181
+ }
182
+ node {
183
+ input: "196"
184
+ input: "318"
185
+ input: "319"
186
+ output: "317"
187
+ name: "Conv_9"
188
+ op_type: "Conv"
189
+ attribute {
190
+ name: "dilations"
191
+ ints: 1
192
+ type: INTS
193
+ }
194
+ attribute {
195
+ name: "group"
196
+ i: 1
197
+ type: INT
198
+ }
199
+ attribute {
200
+ name: "kernel_shape"
201
+ ints: 7
202
+ type: INTS
203
+ }
204
+ attribute {
205
+ name: "pads"
206
+ ints: 3
207
+ ints: 3
208
+ type: INTS
209
+ }
210
+ attribute {
211
+ name: "strides"
212
+ ints: 1
213
+ type: INTS
214
+ }
215
+ }
216
+ node {
217
+ output: "199"
218
+ name: "Constant_10"
219
+ op_type: "Constant"
220
+ attribute {
221
+ name: "value"
222
+ t {
223
+ data_type: 1
224
+ raw_data: "\315\314\314="
225
+ }
226
+ type: TENSOR
227
+ }
228
+ }
229
+ node {
230
+ input: "317"
231
+ input: "199"
232
+ output: "200"
233
+ name: "Mul_11"
234
+ op_type: "Mul"
235
+ }
236
+ node {
237
+ input: "200"
238
+ input: "193"
239
+ output: "201"
240
+ name: "Add_12"
241
+ op_type: "Add"
242
+ }
243
+ node {
244
+ input: "201"
245
+ input: "321"
246
+ input: "322"
247
+ output: "320"
248
+ name: "Conv_13"
249
+ op_type: "Conv"
250
+ attribute {
251
+ name: "dilations"
252
+ ints: 1
253
+ type: INTS
254
+ }
255
+ attribute {
256
+ name: "group"
257
+ i: 1
258
+ type: INT
259
+ }
260
+ attribute {
261
+ name: "kernel_shape"
262
+ ints: 7
263
+ type: INTS
264
+ }
265
+ attribute {
266
+ name: "pads"
267
+ ints: 3
268
+ ints: 3
269
+ type: INTS
270
+ }
271
+ attribute {
272
+ name: "strides"
273
+ ints: 1
274
+ type: INTS
275
+ }
276
+ }
277
+ node {
278
+ input: "320"
279
+ output: "204"
280
+ name: "Relu_14"
281
+ op_type: "Relu"
282
+ }
283
+ node {
284
+ input: "204"
285
+ input: "324"
286
+ input: "325"
287
+ output: "323"
288
+ name: "Conv_15"
289
+ op_type: "Conv"
290
+ attribute {
291
+ name: "dilations"
292
+ ints: 1
293
+ type: INTS
294
+ }
295
+ attribute {
296
+ name: "group"
297
+ i: 1
298
+ type: INT
299
+ }
300
+ attribute {
301
+ name: "kernel_shape"
302
+ ints: 7
303
+ type: INTS
304
+ }
305
+ attribute {
306
+ name: "pads"
307
+ ints: 3
308
+ ints: 3
309
+ type: INTS
310
+ }
311
+ attribute {
312
+ name: "strides"
313
+ ints: 1
314
+ type: INTS
315
+ }
316
+ }
317
+ node {
318
+ output: "207"
319
+ name: "Constant_16"
320
+ op_type: "Constant"
321
+ attribute {
322
+ name: "value"
323
+ t {
324
+ data_type: 1
325
+ raw_data: "\315\314\314="
326
+ }
327
+ type: TENSOR
328
+ }
329
+ }
330
+ node {
331
+ input: "323"
332
+ input: "207"
333
+ output: "208"
334
+ name: "Mul_17"
335
+ op_type: "Mul"
336
+ }
337
+ node {
338
+ input: "208"
339
+ input: "201"
340
+ output: "209"
341
+ name: "Add_18"
342
+ op_type: "Add"
343
+ }
344
+ node {
345
+ input: "209"
346
+ input: "327"
347
+ input: "328"
348
+ output: "326"
349
+ name: "Conv_19"
350
+ op_type: "Conv"
351
+ attribute {
352
+ name: "dilations"
353
+ ints: 1
354
+ type: INTS
355
+ }
356
+ attribute {
357
+ name: "group"
358
+ i: 1
359
+ type: INT
360
+ }
361
+ attribute {
362
+ name: "kernel_shape"
363
+ ints: 7
364
+ type: INTS
365
+ }
366
+ attribute {
367
+ name: "pads"
368
+ ints: 3
369
+ ints: 3
370
+ type: INTS
371
+ }
372
+ attribute {
373
+ name: "strides"
374
+ ints: 1
375
+ type: INTS
376
+ }
377
+ }
378
+ node {
379
+ input: "326"
380
+ output: "212"
381
+ name: "Relu_20"
382
+ op_type: "Relu"
383
+ }
384
+ node {
385
+ input: "212"
386
+ input: "330"
387
+ input: "331"
388
+ output: "329"
389
+ name: "Conv_21"
390
+ op_type: "Conv"
391
+ attribute {
392
+ name: "dilations"
393
+ ints: 1
394
+ type: INTS
395
+ }
396
+ attribute {
397
+ name: "group"
398
+ i: 1
399
+ type: INT
400
+ }
401
+ attribute {
402
+ name: "kernel_shape"
403
+ ints: 7
404
+ type: INTS
405
+ }
406
+ attribute {
407
+ name: "pads"
408
+ ints: 3
409
+ ints: 3
410
+ type: INTS
411
+ }
412
+ attribute {
413
+ name: "strides"
414
+ ints: 1
415
+ type: INTS
416
+ }
417
+ }
418
+ node {
419
+ output: "215"
420
+ name: "Constant_22"
421
+ op_type: "Constant"
422
+ attribute {
423
+ name: "value"
424
+ t {
425
+ data_type: 1
426
+ raw_data: "\315\314\314="
427
+ }
428
+ type: TENSOR
429
+ }
430
+ }
431
+ node {
432
+ input: "329"
433
+ input: "215"
434
+ output: "216"
435
+ name: "Mul_23"
436
+ op_type: "Mul"
437
+ }
438
+ node {
439
+ input: "216"
440
+ input: "209"
441
+ output: "217"
442
+ name: "Add_24"
443
+ op_type: "Add"
444
+ }
445
+ node {
446
+ input: "217"
447
+ input: "333"
448
+ input: "334"
449
+ output: "332"
450
+ name: "Conv_25"
451
+ op_type: "Conv"
452
+ attribute {
453
+ name: "dilations"
454
+ ints: 1
455
+ type: INTS
456
+ }
457
+ attribute {
458
+ name: "group"
459
+ i: 1
460
+ type: INT
461
+ }
462
+ attribute {
463
+ name: "kernel_shape"
464
+ ints: 7
465
+ type: INTS
466
+ }
467
+ attribute {
468
+ name: "pads"
469
+ ints: 3
470
+ ints: 3
471
+ type: INTS
472
+ }
473
+ attribute {
474
+ name: "strides"
475
+ ints: 1
476
+ type: INTS
477
+ }
478
+ }
479
+ node {
480
+ input: "332"
481
+ output: "220"
482
+ name: "Relu_26"
483
+ op_type: "Relu"
484
+ }
485
+ node {
486
+ input: "220"
487
+ input: "336"
488
+ input: "337"
489
+ output: "335"
490
+ name: "Conv_27"
491
+ op_type: "Conv"
492
+ attribute {
493
+ name: "dilations"
494
+ ints: 1
495
+ type: INTS
496
+ }
497
+ attribute {
498
+ name: "group"
499
+ i: 1
500
+ type: INT
501
+ }
502
+ attribute {
503
+ name: "kernel_shape"
504
+ ints: 7
505
+ type: INTS
506
+ }
507
+ attribute {
508
+ name: "pads"
509
+ ints: 3
510
+ ints: 3
511
+ type: INTS
512
+ }
513
+ attribute {
514
+ name: "strides"
515
+ ints: 1
516
+ type: INTS
517
+ }
518
+ }
519
+ node {
520
+ output: "223"
521
+ name: "Constant_28"
522
+ op_type: "Constant"
523
+ attribute {
524
+ name: "value"
525
+ t {
526
+ data_type: 1
527
+ raw_data: "\315\314\314="
528
+ }
529
+ type: TENSOR
530
+ }
531
+ }
532
+ node {
533
+ input: "335"
534
+ input: "223"
535
+ output: "224"
536
+ name: "Mul_29"
537
+ op_type: "Mul"
538
+ }
539
+ node {
540
+ input: "224"
541
+ input: "217"
542
+ output: "225"
543
+ name: "Add_30"
544
+ op_type: "Add"
545
+ }
546
+ node {
547
+ input: "225"
548
+ input: "339"
549
+ input: "340"
550
+ output: "338"
551
+ name: "Conv_31"
552
+ op_type: "Conv"
553
+ attribute {
554
+ name: "dilations"
555
+ ints: 1
556
+ type: INTS
557
+ }
558
+ attribute {
559
+ name: "group"
560
+ i: 1
561
+ type: INT
562
+ }
563
+ attribute {
564
+ name: "kernel_shape"
565
+ ints: 7
566
+ type: INTS
567
+ }
568
+ attribute {
569
+ name: "pads"
570
+ ints: 3
571
+ ints: 3
572
+ type: INTS
573
+ }
574
+ attribute {
575
+ name: "strides"
576
+ ints: 1
577
+ type: INTS
578
+ }
579
+ }
580
+ node {
581
+ input: "338"
582
+ output: "228"
583
+ name: "Relu_32"
584
+ op_type: "Relu"
585
+ }
586
+ node {
587
+ input: "228"
588
+ input: "342"
589
+ input: "343"
590
+ output: "341"
591
+ name: "Conv_33"
592
+ op_type: "Conv"
593
+ attribute {
594
+ name: "dilations"
595
+ ints: 1
596
+ type: INTS
597
+ }
598
+ attribute {
599
+ name: "group"
600
+ i: 1
601
+ type: INT
602
+ }
603
+ attribute {
604
+ name: "kernel_shape"
605
+ ints: 7
606
+ type: INTS
607
+ }
608
+ attribute {
609
+ name: "pads"
610
+ ints: 3
611
+ ints: 3
612
+ type: INTS
613
+ }
614
+ attribute {
615
+ name: "strides"
616
+ ints: 1
617
+ type: INTS
618
+ }
619
+ }
620
+ node {
621
+ output: "231"
622
+ name: "Constant_34"
623
+ op_type: "Constant"
624
+ attribute {
625
+ name: "value"
626
+ t {
627
+ data_type: 1
628
+ raw_data: "\315\314\314="
629
+ }
630
+ type: TENSOR
631
+ }
632
+ }
633
+ node {
634
+ input: "341"
635
+ input: "231"
636
+ output: "232"
637
+ name: "Mul_35"
638
+ op_type: "Mul"
639
+ }
640
+ node {
641
+ input: "232"
642
+ input: "225"
643
+ output: "233"
644
+ name: "Add_36"
645
+ op_type: "Add"
646
+ }
647
+ node {
648
+ input: "233"
649
+ input: "345"
650
+ input: "346"
651
+ output: "344"
652
+ name: "Conv_37"
653
+ op_type: "Conv"
654
+ attribute {
655
+ name: "dilations"
656
+ ints: 1
657
+ type: INTS
658
+ }
659
+ attribute {
660
+ name: "group"
661
+ i: 1
662
+ type: INT
663
+ }
664
+ attribute {
665
+ name: "kernel_shape"
666
+ ints: 7
667
+ type: INTS
668
+ }
669
+ attribute {
670
+ name: "pads"
671
+ ints: 3
672
+ ints: 3
673
+ type: INTS
674
+ }
675
+ attribute {
676
+ name: "strides"
677
+ ints: 1
678
+ type: INTS
679
+ }
680
+ }
681
+ node {
682
+ input: "344"
683
+ output: "236"
684
+ name: "Relu_38"
685
+ op_type: "Relu"
686
+ }
687
+ node {
688
+ input: "236"
689
+ input: "348"
690
+ input: "349"
691
+ output: "347"
692
+ name: "Conv_39"
693
+ op_type: "Conv"
694
+ attribute {
695
+ name: "dilations"
696
+ ints: 1
697
+ type: INTS
698
+ }
699
+ attribute {
700
+ name: "group"
701
+ i: 1
702
+ type: INT
703
+ }
704
+ attribute {
705
+ name: "kernel_shape"
706
+ ints: 7
707
+ type: INTS
708
+ }
709
+ attribute {
710
+ name: "pads"
711
+ ints: 3
712
+ ints: 3
713
+ type: INTS
714
+ }
715
+ attribute {
716
+ name: "strides"
717
+ ints: 1
718
+ type: INTS
719
+ }
720
+ }
721
+ node {
722
+ output: "239"
723
+ name: "Constant_40"
724
+ op_type: "Constant"
725
+ attribute {
726
+ name: "value"
727
+ t {
728
+ data_type: 1
729
+ raw_data: "\315\314\314="
730
+ }
731
+ type: TENSOR
732
+ }
733
+ }
734
+ node {
735
+ input: "347"
736
+ input: "239"
737
+ output: "240"
738
+ name: "Mul_41"
739
+ op_type: "Mul"
740
+ }
741
+ node {
742
+ input: "240"
743
+ input: "233"
744
+ output: "241"
745
+ name: "Add_42"
746
+ op_type: "Add"
747
+ }
748
+ node {
749
+ input: "241"
750
+ input: "351"
751
+ input: "352"
752
+ output: "350"
753
+ name: "Conv_43"
754
+ op_type: "Conv"
755
+ attribute {
756
+ name: "dilations"
757
+ ints: 1
758
+ type: INTS
759
+ }
760
+ attribute {
761
+ name: "group"
762
+ i: 1
763
+ type: INT
764
+ }
765
+ attribute {
766
+ name: "kernel_shape"
767
+ ints: 7
768
+ type: INTS
769
+ }
770
+ attribute {
771
+ name: "pads"
772
+ ints: 3
773
+ ints: 3
774
+ type: INTS
775
+ }
776
+ attribute {
777
+ name: "strides"
778
+ ints: 1
779
+ type: INTS
780
+ }
781
+ }
782
+ node {
783
+ input: "350"
784
+ output: "244"
785
+ name: "Relu_44"
786
+ op_type: "Relu"
787
+ }
788
+ node {
789
+ input: "244"
790
+ input: "354"
791
+ input: "355"
792
+ output: "353"
793
+ name: "Conv_45"
794
+ op_type: "Conv"
795
+ attribute {
796
+ name: "dilations"
797
+ ints: 1
798
+ type: INTS
799
+ }
800
+ attribute {
801
+ name: "group"
802
+ i: 1
803
+ type: INT
804
+ }
805
+ attribute {
806
+ name: "kernel_shape"
807
+ ints: 7
808
+ type: INTS
809
+ }
810
+ attribute {
811
+ name: "pads"
812
+ ints: 3
813
+ ints: 3
814
+ type: INTS
815
+ }
816
+ attribute {
817
+ name: "strides"
818
+ ints: 1
819
+ type: INTS
820
+ }
821
+ }
822
+ node {
823
+ output: "247"
824
+ name: "Constant_46"
825
+ op_type: "Constant"
826
+ attribute {
827
+ name: "value"
828
+ t {
829
+ data_type: 1
830
+ raw_data: "\315\314\314="
831
+ }
832
+ type: TENSOR
833
+ }
834
+ }
835
+ node {
836
+ input: "353"
837
+ input: "247"
838
+ output: "248"
839
+ name: "Mul_47"
840
+ op_type: "Mul"
841
+ }
842
+ node {
843
+ input: "248"
844
+ input: "241"
845
+ output: "249"
846
+ name: "Add_48"
847
+ op_type: "Add"
848
+ }
849
+ node {
850
+ input: "249"
851
+ input: "357"
852
+ input: "358"
853
+ output: "356"
854
+ name: "Conv_49"
855
+ op_type: "Conv"
856
+ attribute {
857
+ name: "dilations"
858
+ ints: 1
859
+ type: INTS
860
+ }
861
+ attribute {
862
+ name: "group"
863
+ i: 1
864
+ type: INT
865
+ }
866
+ attribute {
867
+ name: "kernel_shape"
868
+ ints: 7
869
+ type: INTS
870
+ }
871
+ attribute {
872
+ name: "pads"
873
+ ints: 3
874
+ ints: 3
875
+ type: INTS
876
+ }
877
+ attribute {
878
+ name: "strides"
879
+ ints: 1
880
+ type: INTS
881
+ }
882
+ }
883
+ node {
884
+ input: "356"
885
+ output: "252"
886
+ name: "Relu_50"
887
+ op_type: "Relu"
888
+ }
889
+ node {
890
+ input: "252"
891
+ input: "360"
892
+ input: "361"
893
+ output: "359"
894
+ name: "Conv_51"
895
+ op_type: "Conv"
896
+ attribute {
897
+ name: "dilations"
898
+ ints: 1
899
+ type: INTS
900
+ }
901
+ attribute {
902
+ name: "group"
903
+ i: 1
904
+ type: INT
905
+ }
906
+ attribute {
907
+ name: "kernel_shape"
908
+ ints: 7
909
+ type: INTS
910
+ }
911
+ attribute {
912
+ name: "pads"
913
+ ints: 3
914
+ ints: 3
915
+ type: INTS
916
+ }
917
+ attribute {
918
+ name: "strides"
919
+ ints: 1
920
+ type: INTS
921
+ }
922
+ }
923
+ node {
924
+ output: "255"
925
+ name: "Constant_52"
926
+ op_type: "Constant"
927
+ attribute {
928
+ name: "value"
929
+ t {
930
+ data_type: 1
931
+ raw_data: "\315\314\314="
932
+ }
933
+ type: TENSOR
934
+ }
935
+ }
936
+ node {
937
+ input: "359"
938
+ input: "255"
939
+ output: "256"
940
+ name: "Mul_53"
941
+ op_type: "Mul"
942
+ }
943
+ node {
944
+ input: "256"
945
+ input: "249"
946
+ output: "257"
947
+ name: "Add_54"
948
+ op_type: "Add"
949
+ }
950
+ node {
951
+ input: "257"
952
+ input: "363"
953
+ input: "364"
954
+ output: "362"
955
+ name: "Conv_55"
956
+ op_type: "Conv"
957
+ attribute {
958
+ name: "dilations"
959
+ ints: 1
960
+ type: INTS
961
+ }
962
+ attribute {
963
+ name: "group"
964
+ i: 1
965
+ type: INT
966
+ }
967
+ attribute {
968
+ name: "kernel_shape"
969
+ ints: 7
970
+ type: INTS
971
+ }
972
+ attribute {
973
+ name: "pads"
974
+ ints: 3
975
+ ints: 3
976
+ type: INTS
977
+ }
978
+ attribute {
979
+ name: "strides"
980
+ ints: 1
981
+ type: INTS
982
+ }
983
+ }
984
+ node {
985
+ input: "362"
986
+ output: "260"
987
+ name: "Relu_56"
988
+ op_type: "Relu"
989
+ }
990
+ node {
991
+ input: "260"
992
+ input: "366"
993
+ input: "367"
994
+ output: "365"
995
+ name: "Conv_57"
996
+ op_type: "Conv"
997
+ attribute {
998
+ name: "dilations"
999
+ ints: 1
1000
+ type: INTS
1001
+ }
1002
+ attribute {
1003
+ name: "group"
1004
+ i: 1
1005
+ type: INT
1006
+ }
1007
+ attribute {
1008
+ name: "kernel_shape"
1009
+ ints: 7
1010
+ type: INTS
1011
+ }
1012
+ attribute {
1013
+ name: "pads"
1014
+ ints: 3
1015
+ ints: 3
1016
+ type: INTS
1017
+ }
1018
+ attribute {
1019
+ name: "strides"
1020
+ ints: 1
1021
+ type: INTS
1022
+ }
1023
+ }
1024
+ node {
1025
+ output: "263"
1026
+ name: "Constant_58"
1027
+ op_type: "Constant"
1028
+ attribute {
1029
+ name: "value"
1030
+ t {
1031
+ data_type: 1
1032
+ raw_data: "\315\314\314="
1033
+ }
1034
+ type: TENSOR
1035
+ }
1036
+ }
1037
+ node {
1038
+ input: "365"
1039
+ input: "263"
1040
+ output: "264"
1041
+ name: "Mul_59"
1042
+ op_type: "Mul"
1043
+ }
1044
+ node {
1045
+ input: "264"
1046
+ input: "257"
1047
+ output: "265"
1048
+ name: "Add_60"
1049
+ op_type: "Add"
1050
+ }
1051
+ node {
1052
+ input: "265"
1053
+ input: "369"
1054
+ input: "370"
1055
+ output: "368"
1056
+ name: "Conv_61"
1057
+ op_type: "Conv"
1058
+ attribute {
1059
+ name: "dilations"
1060
+ ints: 1
1061
+ type: INTS
1062
+ }
1063
+ attribute {
1064
+ name: "group"
1065
+ i: 1
1066
+ type: INT
1067
+ }
1068
+ attribute {
1069
+ name: "kernel_shape"
1070
+ ints: 7
1071
+ type: INTS
1072
+ }
1073
+ attribute {
1074
+ name: "pads"
1075
+ ints: 3
1076
+ ints: 3
1077
+ type: INTS
1078
+ }
1079
+ attribute {
1080
+ name: "strides"
1081
+ ints: 1
1082
+ type: INTS
1083
+ }
1084
+ }
1085
+ node {
1086
+ input: "368"
1087
+ output: "268"
1088
+ name: "Relu_62"
1089
+ op_type: "Relu"
1090
+ }
1091
+ node {
1092
+ input: "268"
1093
+ input: "372"
1094
+ input: "373"
1095
+ output: "371"
1096
+ name: "Conv_63"
1097
+ op_type: "Conv"
1098
+ attribute {
1099
+ name: "dilations"
1100
+ ints: 1
1101
+ type: INTS
1102
+ }
1103
+ attribute {
1104
+ name: "group"
1105
+ i: 1
1106
+ type: INT
1107
+ }
1108
+ attribute {
1109
+ name: "kernel_shape"
1110
+ ints: 7
1111
+ type: INTS
1112
+ }
1113
+ attribute {
1114
+ name: "pads"
1115
+ ints: 3
1116
+ ints: 3
1117
+ type: INTS
1118
+ }
1119
+ attribute {
1120
+ name: "strides"
1121
+ ints: 1
1122
+ type: INTS
1123
+ }
1124
+ }
1125
+ node {
1126
+ output: "271"
1127
+ name: "Constant_64"
1128
+ op_type: "Constant"
1129
+ attribute {
1130
+ name: "value"
1131
+ t {
1132
+ data_type: 1
1133
+ raw_data: "\315\314\314="
1134
+ }
1135
+ type: TENSOR
1136
+ }
1137
+ }
1138
+ node {
1139
+ input: "371"
1140
+ input: "271"
1141
+ output: "272"
1142
+ name: "Mul_65"
1143
+ op_type: "Mul"
1144
+ }
1145
+ node {
1146
+ input: "272"
1147
+ input: "265"
1148
+ output: "273"
1149
+ name: "Add_66"
1150
+ op_type: "Add"
1151
+ }
1152
+ node {
1153
+ input: "273"
1154
+ input: "375"
1155
+ input: "376"
1156
+ output: "374"
1157
+ name: "Conv_67"
1158
+ op_type: "Conv"
1159
+ attribute {
1160
+ name: "dilations"
1161
+ ints: 1
1162
+ type: INTS
1163
+ }
1164
+ attribute {
1165
+ name: "group"
1166
+ i: 1
1167
+ type: INT
1168
+ }
1169
+ attribute {
1170
+ name: "kernel_shape"
1171
+ ints: 7
1172
+ type: INTS
1173
+ }
1174
+ attribute {
1175
+ name: "pads"
1176
+ ints: 3
1177
+ ints: 3
1178
+ type: INTS
1179
+ }
1180
+ attribute {
1181
+ name: "strides"
1182
+ ints: 1
1183
+ type: INTS
1184
+ }
1185
+ }
1186
+ node {
1187
+ input: "374"
1188
+ output: "276"
1189
+ name: "Relu_68"
1190
+ op_type: "Relu"
1191
+ }
1192
+ node {
1193
+ input: "276"
1194
+ input: "378"
1195
+ input: "379"
1196
+ output: "377"
1197
+ name: "Conv_69"
1198
+ op_type: "Conv"
1199
+ attribute {
1200
+ name: "dilations"
1201
+ ints: 1
1202
+ type: INTS
1203
+ }
1204
+ attribute {
1205
+ name: "group"
1206
+ i: 1
1207
+ type: INT
1208
+ }
1209
+ attribute {
1210
+ name: "kernel_shape"
1211
+ ints: 7
1212
+ type: INTS
1213
+ }
1214
+ attribute {
1215
+ name: "pads"
1216
+ ints: 3
1217
+ ints: 3
1218
+ type: INTS
1219
+ }
1220
+ attribute {
1221
+ name: "strides"
1222
+ ints: 1
1223
+ type: INTS
1224
+ }
1225
+ }
1226
+ node {
1227
+ output: "279"
1228
+ name: "Constant_70"
1229
+ op_type: "Constant"
1230
+ attribute {
1231
+ name: "value"
1232
+ t {
1233
+ data_type: 1
1234
+ raw_data: "\315\314\314="
1235
+ }
1236
+ type: TENSOR
1237
+ }
1238
+ }
1239
+ node {
1240
+ input: "377"
1241
+ input: "279"
1242
+ output: "280"
1243
+ name: "Mul_71"
1244
+ op_type: "Mul"
1245
+ }
1246
+ node {
1247
+ input: "280"
1248
+ input: "273"
1249
+ output: "281"
1250
+ name: "Add_72"
1251
+ op_type: "Add"
1252
+ }
1253
+ node {
1254
+ input: "281"
1255
+ input: "381"
1256
+ input: "382"
1257
+ output: "380"
1258
+ name: "Conv_73"
1259
+ op_type: "Conv"
1260
+ attribute {
1261
+ name: "dilations"
1262
+ ints: 1
1263
+ type: INTS
1264
+ }
1265
+ attribute {
1266
+ name: "group"
1267
+ i: 1
1268
+ type: INT
1269
+ }
1270
+ attribute {
1271
+ name: "kernel_shape"
1272
+ ints: 7
1273
+ type: INTS
1274
+ }
1275
+ attribute {
1276
+ name: "pads"
1277
+ ints: 3
1278
+ ints: 3
1279
+ type: INTS
1280
+ }
1281
+ attribute {
1282
+ name: "strides"
1283
+ ints: 1
1284
+ type: INTS
1285
+ }
1286
+ }
1287
+ node {
1288
+ input: "380"
1289
+ output: "284"
1290
+ name: "Relu_74"
1291
+ op_type: "Relu"
1292
+ }
1293
+ node {
1294
+ input: "284"
1295
+ input: "384"
1296
+ input: "385"
1297
+ output: "383"
1298
+ name: "Conv_75"
1299
+ op_type: "Conv"
1300
+ attribute {
1301
+ name: "dilations"
1302
+ ints: 1
1303
+ type: INTS
1304
+ }
1305
+ attribute {
1306
+ name: "group"
1307
+ i: 1
1308
+ type: INT
1309
+ }
1310
+ attribute {
1311
+ name: "kernel_shape"
1312
+ ints: 7
1313
+ type: INTS
1314
+ }
1315
+ attribute {
1316
+ name: "pads"
1317
+ ints: 3
1318
+ ints: 3
1319
+ type: INTS
1320
+ }
1321
+ attribute {
1322
+ name: "strides"
1323
+ ints: 1
1324
+ type: INTS
1325
+ }
1326
+ }
1327
+ node {
1328
+ output: "287"
1329
+ name: "Constant_76"
1330
+ op_type: "Constant"
1331
+ attribute {
1332
+ name: "value"
1333
+ t {
1334
+ data_type: 1
1335
+ raw_data: "\315\314\314="
1336
+ }
1337
+ type: TENSOR
1338
+ }
1339
+ }
1340
+ node {
1341
+ input: "383"
1342
+ input: "287"
1343
+ output: "288"
1344
+ name: "Mul_77"
1345
+ op_type: "Mul"
1346
+ }
1347
+ node {
1348
+ input: "288"
1349
+ input: "281"
1350
+ output: "289"
1351
+ name: "Add_78"
1352
+ op_type: "Add"
1353
+ }
1354
+ node {
1355
+ input: "289"
1356
+ input: "387"
1357
+ input: "388"
1358
+ output: "386"
1359
+ name: "Conv_79"
1360
+ op_type: "Conv"
1361
+ attribute {
1362
+ name: "dilations"
1363
+ ints: 1
1364
+ type: INTS
1365
+ }
1366
+ attribute {
1367
+ name: "group"
1368
+ i: 1
1369
+ type: INT
1370
+ }
1371
+ attribute {
1372
+ name: "kernel_shape"
1373
+ ints: 7
1374
+ type: INTS
1375
+ }
1376
+ attribute {
1377
+ name: "pads"
1378
+ ints: 3
1379
+ ints: 3
1380
+ type: INTS
1381
+ }
1382
+ attribute {
1383
+ name: "strides"
1384
+ ints: 1
1385
+ type: INTS
1386
+ }
1387
+ }
1388
+ node {
1389
+ input: "386"
1390
+ output: "292"
1391
+ name: "Relu_80"
1392
+ op_type: "Relu"
1393
+ }
1394
+ node {
1395
+ input: "292"
1396
+ input: "390"
1397
+ input: "391"
1398
+ output: "389"
1399
+ name: "Conv_81"
1400
+ op_type: "Conv"
1401
+ attribute {
1402
+ name: "dilations"
1403
+ ints: 1
1404
+ type: INTS
1405
+ }
1406
+ attribute {
1407
+ name: "group"
1408
+ i: 1
1409
+ type: INT
1410
+ }
1411
+ attribute {
1412
+ name: "kernel_shape"
1413
+ ints: 7
1414
+ type: INTS
1415
+ }
1416
+ attribute {
1417
+ name: "pads"
1418
+ ints: 3
1419
+ ints: 3
1420
+ type: INTS
1421
+ }
1422
+ attribute {
1423
+ name: "strides"
1424
+ ints: 1
1425
+ type: INTS
1426
+ }
1427
+ }
1428
+ node {
1429
+ output: "295"
1430
+ name: "Constant_82"
1431
+ op_type: "Constant"
1432
+ attribute {
1433
+ name: "value"
1434
+ t {
1435
+ data_type: 1
1436
+ raw_data: "\315\314\314="
1437
+ }
1438
+ type: TENSOR
1439
+ }
1440
+ }
1441
+ node {
1442
+ input: "389"
1443
+ input: "295"
1444
+ output: "296"
1445
+ name: "Mul_83"
1446
+ op_type: "Mul"
1447
+ }
1448
+ node {
1449
+ input: "296"
1450
+ input: "289"
1451
+ output: "297"
1452
+ name: "Add_84"
1453
+ op_type: "Add"
1454
+ }
1455
+ node {
1456
+ input: "297"
1457
+ input: "393"
1458
+ input: "394"
1459
+ output: "392"
1460
+ name: "Conv_85"
1461
+ op_type: "Conv"
1462
+ attribute {
1463
+ name: "dilations"
1464
+ ints: 1
1465
+ type: INTS
1466
+ }
1467
+ attribute {
1468
+ name: "group"
1469
+ i: 1
1470
+ type: INT
1471
+ }
1472
+ attribute {
1473
+ name: "kernel_shape"
1474
+ ints: 7
1475
+ type: INTS
1476
+ }
1477
+ attribute {
1478
+ name: "pads"
1479
+ ints: 3
1480
+ ints: 3
1481
+ type: INTS
1482
+ }
1483
+ attribute {
1484
+ name: "strides"
1485
+ ints: 1
1486
+ type: INTS
1487
+ }
1488
+ }
1489
+ node {
1490
+ input: "392"
1491
+ output: "300"
1492
+ name: "Relu_86"
1493
+ op_type: "Relu"
1494
+ }
1495
+ node {
1496
+ input: "300"
1497
+ input: "396"
1498
+ input: "397"
1499
+ output: "395"
1500
+ name: "Conv_87"
1501
+ op_type: "Conv"
1502
+ attribute {
1503
+ name: "dilations"
1504
+ ints: 1
1505
+ type: INTS
1506
+ }
1507
+ attribute {
1508
+ name: "group"
1509
+ i: 1
1510
+ type: INT
1511
+ }
1512
+ attribute {
1513
+ name: "kernel_shape"
1514
+ ints: 7
1515
+ type: INTS
1516
+ }
1517
+ attribute {
1518
+ name: "pads"
1519
+ ints: 3
1520
+ ints: 3
1521
+ type: INTS
1522
+ }
1523
+ attribute {
1524
+ name: "strides"
1525
+ ints: 1
1526
+ type: INTS
1527
+ }
1528
+ }
1529
+ node {
1530
+ output: "303"
1531
+ name: "Constant_88"
1532
+ op_type: "Constant"
1533
+ attribute {
1534
+ name: "value"
1535
+ t {
1536
+ data_type: 1
1537
+ raw_data: "\315\314\314="
1538
+ }
1539
+ type: TENSOR
1540
+ }
1541
+ }
1542
+ node {
1543
+ input: "395"
1544
+ input: "303"
1545
+ output: "304"
1546
+ name: "Mul_89"
1547
+ op_type: "Mul"
1548
+ }
1549
+ node {
1550
+ input: "304"
1551
+ input: "297"
1552
+ output: "305"
1553
+ name: "Add_90"
1554
+ op_type: "Add"
1555
+ }
1556
+ node {
1557
+ input: "305"
1558
+ input: "model.16.weight"
1559
+ input: "model.16.bias"
1560
+ output: "306"
1561
+ name: "Conv_91"
1562
+ op_type: "Conv"
1563
+ attribute {
1564
+ name: "dilations"
1565
+ ints: 1
1566
+ type: INTS
1567
+ }
1568
+ attribute {
1569
+ name: "group"
1570
+ i: 1
1571
+ type: INT
1572
+ }
1573
+ attribute {
1574
+ name: "kernel_shape"
1575
+ ints: 1
1576
+ type: INTS
1577
+ }
1578
+ attribute {
1579
+ name: "pads"
1580
+ ints: 0
1581
+ ints: 0
1582
+ type: INTS
1583
+ }
1584
+ attribute {
1585
+ name: "strides"
1586
+ ints: 1
1587
+ type: INTS
1588
+ }
1589
+ }
1590
+ node {
1591
+ input: "306"
1592
+ input: "x"
1593
+ output: "y"
1594
+ name: "Add_92"
1595
+ op_type: "Add"
1596
+ }
1597
+ initializer {
1598
+ dims: 512
1599
+ dims: 512
1600
+ dims: 7
1601
+ data_type: 1
1602
+ name: "309"
1603
+ }
1604
+ initializer {
1605
+ dims: 512
1606
+ data_type: 1
1607
+ name: "310"
1608
+ }
1609
+ initializer {
1610
+ dims: 512
1611
+ dims: 512
1612
+ dims: 7
1613
+ data_type: 1
1614
+ name: "312"
1615
+ }
1616
+ initializer {
1617
+ dims: 512
1618
+ data_type: 1
1619
+ name: "313"
1620
+ }
1621
+ initializer {
1622
+ dims: 512
1623
+ dims: 512
1624
+ dims: 7
1625
+ data_type: 1
1626
+ name: "315"
1627
+ }
1628
+ initializer {
1629
+ dims: 512
1630
+ data_type: 1
1631
+ name: "316"
1632
+ }
1633
+ initializer {
1634
+ dims: 512
1635
+ dims: 512
1636
+ dims: 7
1637
+ data_type: 1
1638
+ name: "318"
1639
+ }
1640
+ initializer {
1641
+ dims: 512
1642
+ data_type: 1
1643
+ name: "319"
1644
+ }
1645
+ initializer {
1646
+ dims: 512
1647
+ dims: 512
1648
+ dims: 7
1649
+ data_type: 1
1650
+ name: "321"
1651
+ }
1652
+ initializer {
1653
+ dims: 512
1654
+ data_type: 1
1655
+ name: "322"
1656
+ }
1657
+ initializer {
1658
+ dims: 512
1659
+ dims: 512
1660
+ dims: 7
1661
+ data_type: 1
1662
+ name: "324"
1663
+ }
1664
+ initializer {
1665
+ dims: 512
1666
+ data_type: 1
1667
+ name: "325"
1668
+ }
1669
+ initializer {
1670
+ dims: 512
1671
+ dims: 512
1672
+ dims: 7
1673
+ data_type: 1
1674
+ name: "327"
1675
+ }
1676
+ initializer {
1677
+ dims: 512
1678
+ data_type: 1
1679
+ name: "328"
1680
+ }
1681
+ initializer {
1682
+ dims: 512
1683
+ dims: 512
1684
+ dims: 7
1685
+ data_type: 1
1686
+ name: "330"
1687
+ }
1688
+ initializer {
1689
+ dims: 512
1690
+ data_type: 1
1691
+ name: "331"
1692
+ }
1693
+ initializer {
1694
+ dims: 512
1695
+ dims: 512
1696
+ dims: 7
1697
+ data_type: 1
1698
+ name: "333"
1699
+ }
1700
+ initializer {
1701
+ dims: 512
1702
+ data_type: 1
1703
+ name: "334"
1704
+ }
1705
+ initializer {
1706
+ dims: 512
1707
+ dims: 512
1708
+ dims: 7
1709
+ data_type: 1
1710
+ name: "336"
1711
+ }
1712
+ initializer {
1713
+ dims: 512
1714
+ data_type: 1
1715
+ name: "337"
1716
+ }
1717
+ initializer {
1718
+ dims: 512
1719
+ dims: 512
1720
+ dims: 7
1721
+ data_type: 1
1722
+ name: "339"
1723
+ }
1724
+ initializer {
1725
+ dims: 512
1726
+ data_type: 1
1727
+ name: "340"
1728
+ }
1729
+ initializer {
1730
+ dims: 512
1731
+ dims: 512
1732
+ dims: 7
1733
+ data_type: 1
1734
+ name: "342"
1735
+ }
1736
+ initializer {
1737
+ dims: 512
1738
+ data_type: 1
1739
+ name: "343"
1740
+ }
1741
+ initializer {
1742
+ dims: 512
1743
+ dims: 512
1744
+ dims: 7
1745
+ data_type: 1
1746
+ name: "345"
1747
+ }
1748
+ initializer {
1749
+ dims: 512
1750
+ data_type: 1
1751
+ name: "346"
1752
+ }
1753
+ initializer {
1754
+ dims: 512
1755
+ dims: 512
1756
+ dims: 7
1757
+ data_type: 1
1758
+ name: "348"
1759
+ }
1760
+ initializer {
1761
+ dims: 512
1762
+ data_type: 1
1763
+ name: "349"
1764
+ }
1765
+ initializer {
1766
+ dims: 512
1767
+ dims: 512
1768
+ dims: 7
1769
+ data_type: 1
1770
+ name: "351"
1771
+ }
1772
+ initializer {
1773
+ dims: 512
1774
+ data_type: 1
1775
+ name: "352"
1776
+ }
1777
+ initializer {
1778
+ dims: 512
1779
+ dims: 512
1780
+ dims: 7
1781
+ data_type: 1
1782
+ name: "354"
1783
+ }
1784
+ initializer {
1785
+ dims: 512
1786
+ data_type: 1
1787
+ name: "355"
1788
+ }
1789
+ initializer {
1790
+ dims: 512
1791
+ dims: 512
1792
+ dims: 7
1793
+ data_type: 1
1794
+ name: "357"
1795
+ }
1796
+ initializer {
1797
+ dims: 512
1798
+ data_type: 1
1799
+ name: "358"
1800
+ }
1801
+ initializer {
1802
+ dims: 512
1803
+ dims: 512
1804
+ dims: 7
1805
+ data_type: 1
1806
+ name: "360"
1807
+ }
1808
+ initializer {
1809
+ dims: 512
1810
+ data_type: 1
1811
+ name: "361"
1812
+ }
1813
+ initializer {
1814
+ dims: 512
1815
+ dims: 512
1816
+ dims: 7
1817
+ data_type: 1
1818
+ name: "363"
1819
+ }
1820
+ initializer {
1821
+ dims: 512
1822
+ data_type: 1
1823
+ name: "364"
1824
+ }
1825
+ initializer {
1826
+ dims: 512
1827
+ dims: 512
1828
+ dims: 7
1829
+ data_type: 1
1830
+ name: "366"
1831
+ }
1832
+ initializer {
1833
+ dims: 512
1834
+ data_type: 1
1835
+ name: "367"
1836
+ }
1837
+ initializer {
1838
+ dims: 512
1839
+ dims: 512
1840
+ dims: 7
1841
+ data_type: 1
1842
+ name: "369"
1843
+ }
1844
+ initializer {
1845
+ dims: 512
1846
+ data_type: 1
1847
+ name: "370"
1848
+ }
1849
+ initializer {
1850
+ dims: 512
1851
+ dims: 512
1852
+ dims: 7
1853
+ data_type: 1
1854
+ name: "372"
1855
+ }
1856
+ initializer {
1857
+ dims: 512
1858
+ data_type: 1
1859
+ name: "373"
1860
+ }
1861
+ initializer {
1862
+ dims: 512
1863
+ dims: 512
1864
+ dims: 7
1865
+ data_type: 1
1866
+ name: "375"
1867
+ }
1868
+ initializer {
1869
+ dims: 512
1870
+ data_type: 1
1871
+ name: "376"
1872
+ }
1873
+ initializer {
1874
+ dims: 512
1875
+ dims: 512
1876
+ dims: 7
1877
+ data_type: 1
1878
+ name: "378"
1879
+ }
1880
+ initializer {
1881
+ dims: 512
1882
+ data_type: 1
1883
+ name: "379"
1884
+ }
1885
+ initializer {
1886
+ dims: 512
1887
+ dims: 512
1888
+ dims: 7
1889
+ data_type: 1
1890
+ name: "381"
1891
+ }
1892
+ initializer {
1893
+ dims: 512
1894
+ data_type: 1
1895
+ name: "382"
1896
+ }
1897
+ initializer {
1898
+ dims: 512
1899
+ dims: 512
1900
+ dims: 7
1901
+ data_type: 1
1902
+ name: "384"
1903
+ }
1904
+ initializer {
1905
+ dims: 512
1906
+ data_type: 1
1907
+ name: "385"
1908
+ }
1909
+ initializer {
1910
+ dims: 512
1911
+ dims: 512
1912
+ dims: 7
1913
+ data_type: 1
1914
+ name: "387"
1915
+ }
1916
+ initializer {
1917
+ dims: 512
1918
+ data_type: 1
1919
+ name: "388"
1920
+ }
1921
+ initializer {
1922
+ dims: 512
1923
+ dims: 512
1924
+ dims: 7
1925
+ data_type: 1
1926
+ name: "390"
1927
+ }
1928
+ initializer {
1929
+ dims: 512
1930
+ data_type: 1
1931
+ name: "391"
1932
+ }
1933
+ initializer {
1934
+ dims: 512
1935
+ dims: 512
1936
+ dims: 7
1937
+ data_type: 1
1938
+ name: "393"
1939
+ }
1940
+ initializer {
1941
+ dims: 512
1942
+ data_type: 1
1943
+ name: "394"
1944
+ }
1945
+ initializer {
1946
+ dims: 512
1947
+ dims: 512
1948
+ dims: 7
1949
+ data_type: 1
1950
+ name: "396"
1951
+ }
1952
+ initializer {
1953
+ dims: 512
1954
+ data_type: 1
1955
+ name: "397"
1956
+ }
1957
+ initializer {
1958
+ dims: 512
1959
+ data_type: 1
1960
+ name: "model.0.bias"
1961
+ }
1962
+ initializer {
1963
+ dims: 512
1964
+ dims: 2
1965
+ dims: 7
1966
+ data_type: 1
1967
+ name: "model.0.weight"
1968
+ }
1969
+ initializer {
1970
+ dims: 2
1971
+ data_type: 1
1972
+ name: "model.16.bias"
1973
+ raw_data: "\371x*=\246\021\216\273"
1974
+ }
1975
+ initializer {
1976
+ dims: 2
1977
+ dims: 512
1978
+ dims: 1
1979
+ data_type: 1
1980
+ name: "model.16.weight"
1981
+ }
1982
+ input {
1983
+ name: "x"
1984
+ type {
1985
+ tensor_type {
1986
+ elem_type: 1
1987
+ shape {
1988
+ dim {
1989
+ dim_param: "batch_size"
1990
+ }
1991
+ dim {
1992
+ dim_value: 2
1993
+ }
1994
+ dim {
1995
+ dim_value: 8192
1996
+ }
1997
+ }
1998
+ }
1999
+ }
2000
+ }
2001
+ output {
2002
+ name: "y"
2003
+ type {
2004
+ tensor_type {
2005
+ elem_type: 1
2006
+ shape {
2007
+ dim {
2008
+ dim_param: "batch_size"
2009
+ }
2010
+ dim {
2011
+ dim_value: 2
2012
+ }
2013
+ dim {
2014
+ dim_value: 8192
2015
+ }
2016
+ }
2017
+ }
2018
+ }
2019
+ }
2020
+ }
2021
+ opset_import {
2022
+ version: 11
2023
+ }
models/ailia-models/resnetda.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:433e91d11483b61a7c9397da054820ee16fad4e24a7fc5768bccf5cc11e4d67d
3
+ size 220306739
models/ailia-models/resnetda.onnx.prototxt ADDED
@@ -0,0 +1,2023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "model.0.weight"
10
+ input: "model.0.bias"
11
+ output: "65"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 7
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 3
32
+ ints: 3
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 1
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "65"
43
+ input: "model.1.body.0.weight"
44
+ input: "model.1.body.0.bias"
45
+ output: "66"
46
+ name: "Conv_1"
47
+ op_type: "Conv"
48
+ attribute {
49
+ name: "dilations"
50
+ ints: 1
51
+ type: INTS
52
+ }
53
+ attribute {
54
+ name: "group"
55
+ i: 1
56
+ type: INT
57
+ }
58
+ attribute {
59
+ name: "kernel_shape"
60
+ ints: 7
61
+ type: INTS
62
+ }
63
+ attribute {
64
+ name: "pads"
65
+ ints: 3
66
+ ints: 3
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "strides"
71
+ ints: 1
72
+ type: INTS
73
+ }
74
+ }
75
+ node {
76
+ input: "66"
77
+ output: "67"
78
+ name: "Relu_2"
79
+ op_type: "Relu"
80
+ }
81
+ node {
82
+ input: "67"
83
+ input: "model.1.body.2.weight"
84
+ input: "model.1.body.2.bias"
85
+ output: "68"
86
+ name: "Conv_3"
87
+ op_type: "Conv"
88
+ attribute {
89
+ name: "dilations"
90
+ ints: 1
91
+ type: INTS
92
+ }
93
+ attribute {
94
+ name: "group"
95
+ i: 1
96
+ type: INT
97
+ }
98
+ attribute {
99
+ name: "kernel_shape"
100
+ ints: 7
101
+ type: INTS
102
+ }
103
+ attribute {
104
+ name: "pads"
105
+ ints: 3
106
+ ints: 3
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "strides"
111
+ ints: 1
112
+ type: INTS
113
+ }
114
+ }
115
+ node {
116
+ output: "69"
117
+ name: "Constant_4"
118
+ op_type: "Constant"
119
+ attribute {
120
+ name: "value"
121
+ t {
122
+ data_type: 1
123
+ raw_data: "\315\314\314="
124
+ }
125
+ type: TENSOR
126
+ }
127
+ }
128
+ node {
129
+ input: "68"
130
+ input: "69"
131
+ output: "70"
132
+ name: "Mul_5"
133
+ op_type: "Mul"
134
+ }
135
+ node {
136
+ input: "70"
137
+ input: "65"
138
+ output: "71"
139
+ name: "Add_6"
140
+ op_type: "Add"
141
+ }
142
+ node {
143
+ input: "71"
144
+ input: "model.2.body.0.weight"
145
+ input: "model.2.body.0.bias"
146
+ output: "72"
147
+ name: "Conv_7"
148
+ op_type: "Conv"
149
+ attribute {
150
+ name: "dilations"
151
+ ints: 1
152
+ type: INTS
153
+ }
154
+ attribute {
155
+ name: "group"
156
+ i: 1
157
+ type: INT
158
+ }
159
+ attribute {
160
+ name: "kernel_shape"
161
+ ints: 7
162
+ type: INTS
163
+ }
164
+ attribute {
165
+ name: "pads"
166
+ ints: 3
167
+ ints: 3
168
+ type: INTS
169
+ }
170
+ attribute {
171
+ name: "strides"
172
+ ints: 1
173
+ type: INTS
174
+ }
175
+ }
176
+ node {
177
+ input: "72"
178
+ output: "73"
179
+ name: "Relu_8"
180
+ op_type: "Relu"
181
+ }
182
+ node {
183
+ input: "73"
184
+ input: "model.2.body.2.weight"
185
+ input: "model.2.body.2.bias"
186
+ output: "74"
187
+ name: "Conv_9"
188
+ op_type: "Conv"
189
+ attribute {
190
+ name: "dilations"
191
+ ints: 1
192
+ type: INTS
193
+ }
194
+ attribute {
195
+ name: "group"
196
+ i: 1
197
+ type: INT
198
+ }
199
+ attribute {
200
+ name: "kernel_shape"
201
+ ints: 7
202
+ type: INTS
203
+ }
204
+ attribute {
205
+ name: "pads"
206
+ ints: 3
207
+ ints: 3
208
+ type: INTS
209
+ }
210
+ attribute {
211
+ name: "strides"
212
+ ints: 1
213
+ type: INTS
214
+ }
215
+ }
216
+ node {
217
+ output: "75"
218
+ name: "Constant_10"
219
+ op_type: "Constant"
220
+ attribute {
221
+ name: "value"
222
+ t {
223
+ data_type: 1
224
+ raw_data: "\315\314\314="
225
+ }
226
+ type: TENSOR
227
+ }
228
+ }
229
+ node {
230
+ input: "74"
231
+ input: "75"
232
+ output: "76"
233
+ name: "Mul_11"
234
+ op_type: "Mul"
235
+ }
236
+ node {
237
+ input: "76"
238
+ input: "71"
239
+ output: "77"
240
+ name: "Add_12"
241
+ op_type: "Add"
242
+ }
243
+ node {
244
+ input: "77"
245
+ input: "model.3.body.0.weight"
246
+ input: "model.3.body.0.bias"
247
+ output: "78"
248
+ name: "Conv_13"
249
+ op_type: "Conv"
250
+ attribute {
251
+ name: "dilations"
252
+ ints: 1
253
+ type: INTS
254
+ }
255
+ attribute {
256
+ name: "group"
257
+ i: 1
258
+ type: INT
259
+ }
260
+ attribute {
261
+ name: "kernel_shape"
262
+ ints: 7
263
+ type: INTS
264
+ }
265
+ attribute {
266
+ name: "pads"
267
+ ints: 3
268
+ ints: 3
269
+ type: INTS
270
+ }
271
+ attribute {
272
+ name: "strides"
273
+ ints: 1
274
+ type: INTS
275
+ }
276
+ }
277
+ node {
278
+ input: "78"
279
+ output: "79"
280
+ name: "Relu_14"
281
+ op_type: "Relu"
282
+ }
283
+ node {
284
+ input: "79"
285
+ input: "model.3.body.2.weight"
286
+ input: "model.3.body.2.bias"
287
+ output: "80"
288
+ name: "Conv_15"
289
+ op_type: "Conv"
290
+ attribute {
291
+ name: "dilations"
292
+ ints: 1
293
+ type: INTS
294
+ }
295
+ attribute {
296
+ name: "group"
297
+ i: 1
298
+ type: INT
299
+ }
300
+ attribute {
301
+ name: "kernel_shape"
302
+ ints: 7
303
+ type: INTS
304
+ }
305
+ attribute {
306
+ name: "pads"
307
+ ints: 3
308
+ ints: 3
309
+ type: INTS
310
+ }
311
+ attribute {
312
+ name: "strides"
313
+ ints: 1
314
+ type: INTS
315
+ }
316
+ }
317
+ node {
318
+ output: "81"
319
+ name: "Constant_16"
320
+ op_type: "Constant"
321
+ attribute {
322
+ name: "value"
323
+ t {
324
+ data_type: 1
325
+ raw_data: "\315\314\314="
326
+ }
327
+ type: TENSOR
328
+ }
329
+ }
330
+ node {
331
+ input: "80"
332
+ input: "81"
333
+ output: "82"
334
+ name: "Mul_17"
335
+ op_type: "Mul"
336
+ }
337
+ node {
338
+ input: "82"
339
+ input: "77"
340
+ output: "83"
341
+ name: "Add_18"
342
+ op_type: "Add"
343
+ }
344
+ node {
345
+ input: "83"
346
+ input: "model.4.body.0.weight"
347
+ input: "model.4.body.0.bias"
348
+ output: "84"
349
+ name: "Conv_19"
350
+ op_type: "Conv"
351
+ attribute {
352
+ name: "dilations"
353
+ ints: 1
354
+ type: INTS
355
+ }
356
+ attribute {
357
+ name: "group"
358
+ i: 1
359
+ type: INT
360
+ }
361
+ attribute {
362
+ name: "kernel_shape"
363
+ ints: 7
364
+ type: INTS
365
+ }
366
+ attribute {
367
+ name: "pads"
368
+ ints: 3
369
+ ints: 3
370
+ type: INTS
371
+ }
372
+ attribute {
373
+ name: "strides"
374
+ ints: 1
375
+ type: INTS
376
+ }
377
+ }
378
+ node {
379
+ input: "84"
380
+ output: "85"
381
+ name: "Relu_20"
382
+ op_type: "Relu"
383
+ }
384
+ node {
385
+ input: "85"
386
+ input: "model.4.body.2.weight"
387
+ input: "model.4.body.2.bias"
388
+ output: "86"
389
+ name: "Conv_21"
390
+ op_type: "Conv"
391
+ attribute {
392
+ name: "dilations"
393
+ ints: 1
394
+ type: INTS
395
+ }
396
+ attribute {
397
+ name: "group"
398
+ i: 1
399
+ type: INT
400
+ }
401
+ attribute {
402
+ name: "kernel_shape"
403
+ ints: 7
404
+ type: INTS
405
+ }
406
+ attribute {
407
+ name: "pads"
408
+ ints: 3
409
+ ints: 3
410
+ type: INTS
411
+ }
412
+ attribute {
413
+ name: "strides"
414
+ ints: 1
415
+ type: INTS
416
+ }
417
+ }
418
+ node {
419
+ output: "87"
420
+ name: "Constant_22"
421
+ op_type: "Constant"
422
+ attribute {
423
+ name: "value"
424
+ t {
425
+ data_type: 1
426
+ raw_data: "\315\314\314="
427
+ }
428
+ type: TENSOR
429
+ }
430
+ }
431
+ node {
432
+ input: "86"
433
+ input: "87"
434
+ output: "88"
435
+ name: "Mul_23"
436
+ op_type: "Mul"
437
+ }
438
+ node {
439
+ input: "88"
440
+ input: "83"
441
+ output: "89"
442
+ name: "Add_24"
443
+ op_type: "Add"
444
+ }
445
+ node {
446
+ input: "89"
447
+ input: "model.5.body.0.weight"
448
+ input: "model.5.body.0.bias"
449
+ output: "90"
450
+ name: "Conv_25"
451
+ op_type: "Conv"
452
+ attribute {
453
+ name: "dilations"
454
+ ints: 1
455
+ type: INTS
456
+ }
457
+ attribute {
458
+ name: "group"
459
+ i: 1
460
+ type: INT
461
+ }
462
+ attribute {
463
+ name: "kernel_shape"
464
+ ints: 7
465
+ type: INTS
466
+ }
467
+ attribute {
468
+ name: "pads"
469
+ ints: 3
470
+ ints: 3
471
+ type: INTS
472
+ }
473
+ attribute {
474
+ name: "strides"
475
+ ints: 1
476
+ type: INTS
477
+ }
478
+ }
479
+ node {
480
+ input: "90"
481
+ output: "91"
482
+ name: "Relu_26"
483
+ op_type: "Relu"
484
+ }
485
+ node {
486
+ input: "91"
487
+ input: "model.5.body.2.weight"
488
+ input: "model.5.body.2.bias"
489
+ output: "92"
490
+ name: "Conv_27"
491
+ op_type: "Conv"
492
+ attribute {
493
+ name: "dilations"
494
+ ints: 1
495
+ type: INTS
496
+ }
497
+ attribute {
498
+ name: "group"
499
+ i: 1
500
+ type: INT
501
+ }
502
+ attribute {
503
+ name: "kernel_shape"
504
+ ints: 7
505
+ type: INTS
506
+ }
507
+ attribute {
508
+ name: "pads"
509
+ ints: 3
510
+ ints: 3
511
+ type: INTS
512
+ }
513
+ attribute {
514
+ name: "strides"
515
+ ints: 1
516
+ type: INTS
517
+ }
518
+ }
519
+ node {
520
+ output: "93"
521
+ name: "Constant_28"
522
+ op_type: "Constant"
523
+ attribute {
524
+ name: "value"
525
+ t {
526
+ data_type: 1
527
+ raw_data: "\315\314\314="
528
+ }
529
+ type: TENSOR
530
+ }
531
+ }
532
+ node {
533
+ input: "92"
534
+ input: "93"
535
+ output: "94"
536
+ name: "Mul_29"
537
+ op_type: "Mul"
538
+ }
539
+ node {
540
+ input: "94"
541
+ input: "89"
542
+ output: "95"
543
+ name: "Add_30"
544
+ op_type: "Add"
545
+ }
546
+ node {
547
+ input: "95"
548
+ input: "model.6.body.0.weight"
549
+ input: "model.6.body.0.bias"
550
+ output: "96"
551
+ name: "Conv_31"
552
+ op_type: "Conv"
553
+ attribute {
554
+ name: "dilations"
555
+ ints: 1
556
+ type: INTS
557
+ }
558
+ attribute {
559
+ name: "group"
560
+ i: 1
561
+ type: INT
562
+ }
563
+ attribute {
564
+ name: "kernel_shape"
565
+ ints: 7
566
+ type: INTS
567
+ }
568
+ attribute {
569
+ name: "pads"
570
+ ints: 3
571
+ ints: 3
572
+ type: INTS
573
+ }
574
+ attribute {
575
+ name: "strides"
576
+ ints: 1
577
+ type: INTS
578
+ }
579
+ }
580
+ node {
581
+ input: "96"
582
+ output: "97"
583
+ name: "Relu_32"
584
+ op_type: "Relu"
585
+ }
586
+ node {
587
+ input: "97"
588
+ input: "model.6.body.2.weight"
589
+ input: "model.6.body.2.bias"
590
+ output: "98"
591
+ name: "Conv_33"
592
+ op_type: "Conv"
593
+ attribute {
594
+ name: "dilations"
595
+ ints: 1
596
+ type: INTS
597
+ }
598
+ attribute {
599
+ name: "group"
600
+ i: 1
601
+ type: INT
602
+ }
603
+ attribute {
604
+ name: "kernel_shape"
605
+ ints: 7
606
+ type: INTS
607
+ }
608
+ attribute {
609
+ name: "pads"
610
+ ints: 3
611
+ ints: 3
612
+ type: INTS
613
+ }
614
+ attribute {
615
+ name: "strides"
616
+ ints: 1
617
+ type: INTS
618
+ }
619
+ }
620
+ node {
621
+ output: "99"
622
+ name: "Constant_34"
623
+ op_type: "Constant"
624
+ attribute {
625
+ name: "value"
626
+ t {
627
+ data_type: 1
628
+ raw_data: "\315\314\314="
629
+ }
630
+ type: TENSOR
631
+ }
632
+ }
633
+ node {
634
+ input: "98"
635
+ input: "99"
636
+ output: "100"
637
+ name: "Mul_35"
638
+ op_type: "Mul"
639
+ }
640
+ node {
641
+ input: "100"
642
+ input: "95"
643
+ output: "101"
644
+ name: "Add_36"
645
+ op_type: "Add"
646
+ }
647
+ node {
648
+ input: "101"
649
+ input: "model.7.body.0.weight"
650
+ input: "model.7.body.0.bias"
651
+ output: "102"
652
+ name: "Conv_37"
653
+ op_type: "Conv"
654
+ attribute {
655
+ name: "dilations"
656
+ ints: 1
657
+ type: INTS
658
+ }
659
+ attribute {
660
+ name: "group"
661
+ i: 1
662
+ type: INT
663
+ }
664
+ attribute {
665
+ name: "kernel_shape"
666
+ ints: 7
667
+ type: INTS
668
+ }
669
+ attribute {
670
+ name: "pads"
671
+ ints: 3
672
+ ints: 3
673
+ type: INTS
674
+ }
675
+ attribute {
676
+ name: "strides"
677
+ ints: 1
678
+ type: INTS
679
+ }
680
+ }
681
+ node {
682
+ input: "102"
683
+ output: "103"
684
+ name: "Relu_38"
685
+ op_type: "Relu"
686
+ }
687
+ node {
688
+ input: "103"
689
+ input: "model.7.body.2.weight"
690
+ input: "model.7.body.2.bias"
691
+ output: "104"
692
+ name: "Conv_39"
693
+ op_type: "Conv"
694
+ attribute {
695
+ name: "dilations"
696
+ ints: 1
697
+ type: INTS
698
+ }
699
+ attribute {
700
+ name: "group"
701
+ i: 1
702
+ type: INT
703
+ }
704
+ attribute {
705
+ name: "kernel_shape"
706
+ ints: 7
707
+ type: INTS
708
+ }
709
+ attribute {
710
+ name: "pads"
711
+ ints: 3
712
+ ints: 3
713
+ type: INTS
714
+ }
715
+ attribute {
716
+ name: "strides"
717
+ ints: 1
718
+ type: INTS
719
+ }
720
+ }
721
+ node {
722
+ output: "105"
723
+ name: "Constant_40"
724
+ op_type: "Constant"
725
+ attribute {
726
+ name: "value"
727
+ t {
728
+ data_type: 1
729
+ raw_data: "\315\314\314="
730
+ }
731
+ type: TENSOR
732
+ }
733
+ }
734
+ node {
735
+ input: "104"
736
+ input: "105"
737
+ output: "106"
738
+ name: "Mul_41"
739
+ op_type: "Mul"
740
+ }
741
+ node {
742
+ input: "106"
743
+ input: "101"
744
+ output: "107"
745
+ name: "Add_42"
746
+ op_type: "Add"
747
+ }
748
+ node {
749
+ input: "107"
750
+ input: "model.8.body.0.weight"
751
+ input: "model.8.body.0.bias"
752
+ output: "108"
753
+ name: "Conv_43"
754
+ op_type: "Conv"
755
+ attribute {
756
+ name: "dilations"
757
+ ints: 1
758
+ type: INTS
759
+ }
760
+ attribute {
761
+ name: "group"
762
+ i: 1
763
+ type: INT
764
+ }
765
+ attribute {
766
+ name: "kernel_shape"
767
+ ints: 7
768
+ type: INTS
769
+ }
770
+ attribute {
771
+ name: "pads"
772
+ ints: 3
773
+ ints: 3
774
+ type: INTS
775
+ }
776
+ attribute {
777
+ name: "strides"
778
+ ints: 1
779
+ type: INTS
780
+ }
781
+ }
782
+ node {
783
+ input: "108"
784
+ output: "109"
785
+ name: "Relu_44"
786
+ op_type: "Relu"
787
+ }
788
+ node {
789
+ input: "109"
790
+ input: "model.8.body.2.weight"
791
+ input: "model.8.body.2.bias"
792
+ output: "110"
793
+ name: "Conv_45"
794
+ op_type: "Conv"
795
+ attribute {
796
+ name: "dilations"
797
+ ints: 1
798
+ type: INTS
799
+ }
800
+ attribute {
801
+ name: "group"
802
+ i: 1
803
+ type: INT
804
+ }
805
+ attribute {
806
+ name: "kernel_shape"
807
+ ints: 7
808
+ type: INTS
809
+ }
810
+ attribute {
811
+ name: "pads"
812
+ ints: 3
813
+ ints: 3
814
+ type: INTS
815
+ }
816
+ attribute {
817
+ name: "strides"
818
+ ints: 1
819
+ type: INTS
820
+ }
821
+ }
822
+ node {
823
+ output: "111"
824
+ name: "Constant_46"
825
+ op_type: "Constant"
826
+ attribute {
827
+ name: "value"
828
+ t {
829
+ data_type: 1
830
+ raw_data: "\315\314\314="
831
+ }
832
+ type: TENSOR
833
+ }
834
+ }
835
+ node {
836
+ input: "110"
837
+ input: "111"
838
+ output: "112"
839
+ name: "Mul_47"
840
+ op_type: "Mul"
841
+ }
842
+ node {
843
+ input: "112"
844
+ input: "107"
845
+ output: "113"
846
+ name: "Add_48"
847
+ op_type: "Add"
848
+ }
849
+ node {
850
+ input: "113"
851
+ input: "model.9.body.0.weight"
852
+ input: "model.9.body.0.bias"
853
+ output: "114"
854
+ name: "Conv_49"
855
+ op_type: "Conv"
856
+ attribute {
857
+ name: "dilations"
858
+ ints: 1
859
+ type: INTS
860
+ }
861
+ attribute {
862
+ name: "group"
863
+ i: 1
864
+ type: INT
865
+ }
866
+ attribute {
867
+ name: "kernel_shape"
868
+ ints: 7
869
+ type: INTS
870
+ }
871
+ attribute {
872
+ name: "pads"
873
+ ints: 3
874
+ ints: 3
875
+ type: INTS
876
+ }
877
+ attribute {
878
+ name: "strides"
879
+ ints: 1
880
+ type: INTS
881
+ }
882
+ }
883
+ node {
884
+ input: "114"
885
+ output: "115"
886
+ name: "Relu_50"
887
+ op_type: "Relu"
888
+ }
889
+ node {
890
+ input: "115"
891
+ input: "model.9.body.2.weight"
892
+ input: "model.9.body.2.bias"
893
+ output: "116"
894
+ name: "Conv_51"
895
+ op_type: "Conv"
896
+ attribute {
897
+ name: "dilations"
898
+ ints: 1
899
+ type: INTS
900
+ }
901
+ attribute {
902
+ name: "group"
903
+ i: 1
904
+ type: INT
905
+ }
906
+ attribute {
907
+ name: "kernel_shape"
908
+ ints: 7
909
+ type: INTS
910
+ }
911
+ attribute {
912
+ name: "pads"
913
+ ints: 3
914
+ ints: 3
915
+ type: INTS
916
+ }
917
+ attribute {
918
+ name: "strides"
919
+ ints: 1
920
+ type: INTS
921
+ }
922
+ }
923
+ node {
924
+ output: "117"
925
+ name: "Constant_52"
926
+ op_type: "Constant"
927
+ attribute {
928
+ name: "value"
929
+ t {
930
+ data_type: 1
931
+ raw_data: "\315\314\314="
932
+ }
933
+ type: TENSOR
934
+ }
935
+ }
936
+ node {
937
+ input: "116"
938
+ input: "117"
939
+ output: "118"
940
+ name: "Mul_53"
941
+ op_type: "Mul"
942
+ }
943
+ node {
944
+ input: "118"
945
+ input: "113"
946
+ output: "119"
947
+ name: "Add_54"
948
+ op_type: "Add"
949
+ }
950
+ node {
951
+ input: "119"
952
+ input: "model.10.body.0.weight"
953
+ input: "model.10.body.0.bias"
954
+ output: "120"
955
+ name: "Conv_55"
956
+ op_type: "Conv"
957
+ attribute {
958
+ name: "dilations"
959
+ ints: 1
960
+ type: INTS
961
+ }
962
+ attribute {
963
+ name: "group"
964
+ i: 1
965
+ type: INT
966
+ }
967
+ attribute {
968
+ name: "kernel_shape"
969
+ ints: 7
970
+ type: INTS
971
+ }
972
+ attribute {
973
+ name: "pads"
974
+ ints: 3
975
+ ints: 3
976
+ type: INTS
977
+ }
978
+ attribute {
979
+ name: "strides"
980
+ ints: 1
981
+ type: INTS
982
+ }
983
+ }
984
+ node {
985
+ input: "120"
986
+ output: "121"
987
+ name: "Relu_56"
988
+ op_type: "Relu"
989
+ }
990
+ node {
991
+ input: "121"
992
+ input: "model.10.body.2.weight"
993
+ input: "model.10.body.2.bias"
994
+ output: "122"
995
+ name: "Conv_57"
996
+ op_type: "Conv"
997
+ attribute {
998
+ name: "dilations"
999
+ ints: 1
1000
+ type: INTS
1001
+ }
1002
+ attribute {
1003
+ name: "group"
1004
+ i: 1
1005
+ type: INT
1006
+ }
1007
+ attribute {
1008
+ name: "kernel_shape"
1009
+ ints: 7
1010
+ type: INTS
1011
+ }
1012
+ attribute {
1013
+ name: "pads"
1014
+ ints: 3
1015
+ ints: 3
1016
+ type: INTS
1017
+ }
1018
+ attribute {
1019
+ name: "strides"
1020
+ ints: 1
1021
+ type: INTS
1022
+ }
1023
+ }
1024
+ node {
1025
+ output: "123"
1026
+ name: "Constant_58"
1027
+ op_type: "Constant"
1028
+ attribute {
1029
+ name: "value"
1030
+ t {
1031
+ data_type: 1
1032
+ raw_data: "\315\314\314="
1033
+ }
1034
+ type: TENSOR
1035
+ }
1036
+ }
1037
+ node {
1038
+ input: "122"
1039
+ input: "123"
1040
+ output: "124"
1041
+ name: "Mul_59"
1042
+ op_type: "Mul"
1043
+ }
1044
+ node {
1045
+ input: "124"
1046
+ input: "119"
1047
+ output: "125"
1048
+ name: "Add_60"
1049
+ op_type: "Add"
1050
+ }
1051
+ node {
1052
+ input: "125"
1053
+ input: "model.11.body.0.weight"
1054
+ input: "model.11.body.0.bias"
1055
+ output: "126"
1056
+ name: "Conv_61"
1057
+ op_type: "Conv"
1058
+ attribute {
1059
+ name: "dilations"
1060
+ ints: 1
1061
+ type: INTS
1062
+ }
1063
+ attribute {
1064
+ name: "group"
1065
+ i: 1
1066
+ type: INT
1067
+ }
1068
+ attribute {
1069
+ name: "kernel_shape"
1070
+ ints: 7
1071
+ type: INTS
1072
+ }
1073
+ attribute {
1074
+ name: "pads"
1075
+ ints: 3
1076
+ ints: 3
1077
+ type: INTS
1078
+ }
1079
+ attribute {
1080
+ name: "strides"
1081
+ ints: 1
1082
+ type: INTS
1083
+ }
1084
+ }
1085
+ node {
1086
+ input: "126"
1087
+ output: "127"
1088
+ name: "Relu_62"
1089
+ op_type: "Relu"
1090
+ }
1091
+ node {
1092
+ input: "127"
1093
+ input: "model.11.body.2.weight"
1094
+ input: "model.11.body.2.bias"
1095
+ output: "128"
1096
+ name: "Conv_63"
1097
+ op_type: "Conv"
1098
+ attribute {
1099
+ name: "dilations"
1100
+ ints: 1
1101
+ type: INTS
1102
+ }
1103
+ attribute {
1104
+ name: "group"
1105
+ i: 1
1106
+ type: INT
1107
+ }
1108
+ attribute {
1109
+ name: "kernel_shape"
1110
+ ints: 7
1111
+ type: INTS
1112
+ }
1113
+ attribute {
1114
+ name: "pads"
1115
+ ints: 3
1116
+ ints: 3
1117
+ type: INTS
1118
+ }
1119
+ attribute {
1120
+ name: "strides"
1121
+ ints: 1
1122
+ type: INTS
1123
+ }
1124
+ }
1125
+ node {
1126
+ output: "129"
1127
+ name: "Constant_64"
1128
+ op_type: "Constant"
1129
+ attribute {
1130
+ name: "value"
1131
+ t {
1132
+ data_type: 1
1133
+ raw_data: "\315\314\314="
1134
+ }
1135
+ type: TENSOR
1136
+ }
1137
+ }
1138
+ node {
1139
+ input: "128"
1140
+ input: "129"
1141
+ output: "130"
1142
+ name: "Mul_65"
1143
+ op_type: "Mul"
1144
+ }
1145
+ node {
1146
+ input: "130"
1147
+ input: "125"
1148
+ output: "131"
1149
+ name: "Add_66"
1150
+ op_type: "Add"
1151
+ }
1152
+ node {
1153
+ input: "131"
1154
+ input: "model.12.body.0.weight"
1155
+ input: "model.12.body.0.bias"
1156
+ output: "132"
1157
+ name: "Conv_67"
1158
+ op_type: "Conv"
1159
+ attribute {
1160
+ name: "dilations"
1161
+ ints: 1
1162
+ type: INTS
1163
+ }
1164
+ attribute {
1165
+ name: "group"
1166
+ i: 1
1167
+ type: INT
1168
+ }
1169
+ attribute {
1170
+ name: "kernel_shape"
1171
+ ints: 7
1172
+ type: INTS
1173
+ }
1174
+ attribute {
1175
+ name: "pads"
1176
+ ints: 3
1177
+ ints: 3
1178
+ type: INTS
1179
+ }
1180
+ attribute {
1181
+ name: "strides"
1182
+ ints: 1
1183
+ type: INTS
1184
+ }
1185
+ }
1186
+ node {
1187
+ input: "132"
1188
+ output: "133"
1189
+ name: "Relu_68"
1190
+ op_type: "Relu"
1191
+ }
1192
+ node {
1193
+ input: "133"
1194
+ input: "model.12.body.2.weight"
1195
+ input: "model.12.body.2.bias"
1196
+ output: "134"
1197
+ name: "Conv_69"
1198
+ op_type: "Conv"
1199
+ attribute {
1200
+ name: "dilations"
1201
+ ints: 1
1202
+ type: INTS
1203
+ }
1204
+ attribute {
1205
+ name: "group"
1206
+ i: 1
1207
+ type: INT
1208
+ }
1209
+ attribute {
1210
+ name: "kernel_shape"
1211
+ ints: 7
1212
+ type: INTS
1213
+ }
1214
+ attribute {
1215
+ name: "pads"
1216
+ ints: 3
1217
+ ints: 3
1218
+ type: INTS
1219
+ }
1220
+ attribute {
1221
+ name: "strides"
1222
+ ints: 1
1223
+ type: INTS
1224
+ }
1225
+ }
1226
+ node {
1227
+ output: "135"
1228
+ name: "Constant_70"
1229
+ op_type: "Constant"
1230
+ attribute {
1231
+ name: "value"
1232
+ t {
1233
+ data_type: 1
1234
+ raw_data: "\315\314\314="
1235
+ }
1236
+ type: TENSOR
1237
+ }
1238
+ }
1239
+ node {
1240
+ input: "134"
1241
+ input: "135"
1242
+ output: "136"
1243
+ name: "Mul_71"
1244
+ op_type: "Mul"
1245
+ }
1246
+ node {
1247
+ input: "136"
1248
+ input: "131"
1249
+ output: "137"
1250
+ name: "Add_72"
1251
+ op_type: "Add"
1252
+ }
1253
+ node {
1254
+ input: "137"
1255
+ input: "model.13.body.0.weight"
1256
+ input: "model.13.body.0.bias"
1257
+ output: "138"
1258
+ name: "Conv_73"
1259
+ op_type: "Conv"
1260
+ attribute {
1261
+ name: "dilations"
1262
+ ints: 1
1263
+ type: INTS
1264
+ }
1265
+ attribute {
1266
+ name: "group"
1267
+ i: 1
1268
+ type: INT
1269
+ }
1270
+ attribute {
1271
+ name: "kernel_shape"
1272
+ ints: 7
1273
+ type: INTS
1274
+ }
1275
+ attribute {
1276
+ name: "pads"
1277
+ ints: 3
1278
+ ints: 3
1279
+ type: INTS
1280
+ }
1281
+ attribute {
1282
+ name: "strides"
1283
+ ints: 1
1284
+ type: INTS
1285
+ }
1286
+ }
1287
+ node {
1288
+ input: "138"
1289
+ output: "139"
1290
+ name: "Relu_74"
1291
+ op_type: "Relu"
1292
+ }
1293
+ node {
1294
+ input: "139"
1295
+ input: "model.13.body.2.weight"
1296
+ input: "model.13.body.2.bias"
1297
+ output: "140"
1298
+ name: "Conv_75"
1299
+ op_type: "Conv"
1300
+ attribute {
1301
+ name: "dilations"
1302
+ ints: 1
1303
+ type: INTS
1304
+ }
1305
+ attribute {
1306
+ name: "group"
1307
+ i: 1
1308
+ type: INT
1309
+ }
1310
+ attribute {
1311
+ name: "kernel_shape"
1312
+ ints: 7
1313
+ type: INTS
1314
+ }
1315
+ attribute {
1316
+ name: "pads"
1317
+ ints: 3
1318
+ ints: 3
1319
+ type: INTS
1320
+ }
1321
+ attribute {
1322
+ name: "strides"
1323
+ ints: 1
1324
+ type: INTS
1325
+ }
1326
+ }
1327
+ node {
1328
+ output: "141"
1329
+ name: "Constant_76"
1330
+ op_type: "Constant"
1331
+ attribute {
1332
+ name: "value"
1333
+ t {
1334
+ data_type: 1
1335
+ raw_data: "\315\314\314="
1336
+ }
1337
+ type: TENSOR
1338
+ }
1339
+ }
1340
+ node {
1341
+ input: "140"
1342
+ input: "141"
1343
+ output: "142"
1344
+ name: "Mul_77"
1345
+ op_type: "Mul"
1346
+ }
1347
+ node {
1348
+ input: "142"
1349
+ input: "137"
1350
+ output: "143"
1351
+ name: "Add_78"
1352
+ op_type: "Add"
1353
+ }
1354
+ node {
1355
+ input: "143"
1356
+ input: "model.14.body.0.weight"
1357
+ input: "model.14.body.0.bias"
1358
+ output: "144"
1359
+ name: "Conv_79"
1360
+ op_type: "Conv"
1361
+ attribute {
1362
+ name: "dilations"
1363
+ ints: 1
1364
+ type: INTS
1365
+ }
1366
+ attribute {
1367
+ name: "group"
1368
+ i: 1
1369
+ type: INT
1370
+ }
1371
+ attribute {
1372
+ name: "kernel_shape"
1373
+ ints: 7
1374
+ type: INTS
1375
+ }
1376
+ attribute {
1377
+ name: "pads"
1378
+ ints: 3
1379
+ ints: 3
1380
+ type: INTS
1381
+ }
1382
+ attribute {
1383
+ name: "strides"
1384
+ ints: 1
1385
+ type: INTS
1386
+ }
1387
+ }
1388
+ node {
1389
+ input: "144"
1390
+ output: "145"
1391
+ name: "Relu_80"
1392
+ op_type: "Relu"
1393
+ }
1394
+ node {
1395
+ input: "145"
1396
+ input: "model.14.body.2.weight"
1397
+ input: "model.14.body.2.bias"
1398
+ output: "146"
1399
+ name: "Conv_81"
1400
+ op_type: "Conv"
1401
+ attribute {
1402
+ name: "dilations"
1403
+ ints: 1
1404
+ type: INTS
1405
+ }
1406
+ attribute {
1407
+ name: "group"
1408
+ i: 1
1409
+ type: INT
1410
+ }
1411
+ attribute {
1412
+ name: "kernel_shape"
1413
+ ints: 7
1414
+ type: INTS
1415
+ }
1416
+ attribute {
1417
+ name: "pads"
1418
+ ints: 3
1419
+ ints: 3
1420
+ type: INTS
1421
+ }
1422
+ attribute {
1423
+ name: "strides"
1424
+ ints: 1
1425
+ type: INTS
1426
+ }
1427
+ }
1428
+ node {
1429
+ output: "147"
1430
+ name: "Constant_82"
1431
+ op_type: "Constant"
1432
+ attribute {
1433
+ name: "value"
1434
+ t {
1435
+ data_type: 1
1436
+ raw_data: "\315\314\314="
1437
+ }
1438
+ type: TENSOR
1439
+ }
1440
+ }
1441
+ node {
1442
+ input: "146"
1443
+ input: "147"
1444
+ output: "148"
1445
+ name: "Mul_83"
1446
+ op_type: "Mul"
1447
+ }
1448
+ node {
1449
+ input: "148"
1450
+ input: "143"
1451
+ output: "149"
1452
+ name: "Add_84"
1453
+ op_type: "Add"
1454
+ }
1455
+ node {
1456
+ input: "149"
1457
+ input: "model.15.body.0.weight"
1458
+ input: "model.15.body.0.bias"
1459
+ output: "150"
1460
+ name: "Conv_85"
1461
+ op_type: "Conv"
1462
+ attribute {
1463
+ name: "dilations"
1464
+ ints: 1
1465
+ type: INTS
1466
+ }
1467
+ attribute {
1468
+ name: "group"
1469
+ i: 1
1470
+ type: INT
1471
+ }
1472
+ attribute {
1473
+ name: "kernel_shape"
1474
+ ints: 7
1475
+ type: INTS
1476
+ }
1477
+ attribute {
1478
+ name: "pads"
1479
+ ints: 3
1480
+ ints: 3
1481
+ type: INTS
1482
+ }
1483
+ attribute {
1484
+ name: "strides"
1485
+ ints: 1
1486
+ type: INTS
1487
+ }
1488
+ }
1489
+ node {
1490
+ input: "150"
1491
+ output: "151"
1492
+ name: "Relu_86"
1493
+ op_type: "Relu"
1494
+ }
1495
+ node {
1496
+ input: "151"
1497
+ input: "model.15.body.2.weight"
1498
+ input: "model.15.body.2.bias"
1499
+ output: "152"
1500
+ name: "Conv_87"
1501
+ op_type: "Conv"
1502
+ attribute {
1503
+ name: "dilations"
1504
+ ints: 1
1505
+ type: INTS
1506
+ }
1507
+ attribute {
1508
+ name: "group"
1509
+ i: 1
1510
+ type: INT
1511
+ }
1512
+ attribute {
1513
+ name: "kernel_shape"
1514
+ ints: 7
1515
+ type: INTS
1516
+ }
1517
+ attribute {
1518
+ name: "pads"
1519
+ ints: 3
1520
+ ints: 3
1521
+ type: INTS
1522
+ }
1523
+ attribute {
1524
+ name: "strides"
1525
+ ints: 1
1526
+ type: INTS
1527
+ }
1528
+ }
1529
+ node {
1530
+ output: "153"
1531
+ name: "Constant_88"
1532
+ op_type: "Constant"
1533
+ attribute {
1534
+ name: "value"
1535
+ t {
1536
+ data_type: 1
1537
+ raw_data: "\315\314\314="
1538
+ }
1539
+ type: TENSOR
1540
+ }
1541
+ }
1542
+ node {
1543
+ input: "152"
1544
+ input: "153"
1545
+ output: "154"
1546
+ name: "Mul_89"
1547
+ op_type: "Mul"
1548
+ }
1549
+ node {
1550
+ input: "154"
1551
+ input: "149"
1552
+ output: "155"
1553
+ name: "Add_90"
1554
+ op_type: "Add"
1555
+ }
1556
+ node {
1557
+ input: "155"
1558
+ input: "model.16.weight"
1559
+ input: "model.16.bias"
1560
+ output: "156"
1561
+ name: "Conv_91"
1562
+ op_type: "Conv"
1563
+ attribute {
1564
+ name: "dilations"
1565
+ ints: 1
1566
+ type: INTS
1567
+ }
1568
+ attribute {
1569
+ name: "group"
1570
+ i: 1
1571
+ type: INT
1572
+ }
1573
+ attribute {
1574
+ name: "kernel_shape"
1575
+ ints: 1
1576
+ type: INTS
1577
+ }
1578
+ attribute {
1579
+ name: "pads"
1580
+ ints: 0
1581
+ ints: 0
1582
+ type: INTS
1583
+ }
1584
+ attribute {
1585
+ name: "strides"
1586
+ ints: 1
1587
+ type: INTS
1588
+ }
1589
+ }
1590
+ node {
1591
+ input: "156"
1592
+ input: "x"
1593
+ output: "y"
1594
+ name: "Add_92"
1595
+ op_type: "Add"
1596
+ }
1597
+ initializer {
1598
+ dims: 512
1599
+ data_type: 1
1600
+ name: "model.0.bias"
1601
+ }
1602
+ initializer {
1603
+ dims: 512
1604
+ dims: 2
1605
+ dims: 7
1606
+ data_type: 1
1607
+ name: "model.0.weight"
1608
+ }
1609
+ initializer {
1610
+ dims: 512
1611
+ data_type: 1
1612
+ name: "model.1.body.0.bias"
1613
+ }
1614
+ initializer {
1615
+ dims: 512
1616
+ dims: 512
1617
+ dims: 7
1618
+ data_type: 1
1619
+ name: "model.1.body.0.weight"
1620
+ }
1621
+ initializer {
1622
+ dims: 512
1623
+ data_type: 1
1624
+ name: "model.1.body.2.bias"
1625
+ }
1626
+ initializer {
1627
+ dims: 512
1628
+ dims: 512
1629
+ dims: 7
1630
+ data_type: 1
1631
+ name: "model.1.body.2.weight"
1632
+ }
1633
+ initializer {
1634
+ dims: 512
1635
+ data_type: 1
1636
+ name: "model.10.body.0.bias"
1637
+ }
1638
+ initializer {
1639
+ dims: 512
1640
+ dims: 512
1641
+ dims: 7
1642
+ data_type: 1
1643
+ name: "model.10.body.0.weight"
1644
+ }
1645
+ initializer {
1646
+ dims: 512
1647
+ data_type: 1
1648
+ name: "model.10.body.2.bias"
1649
+ }
1650
+ initializer {
1651
+ dims: 512
1652
+ dims: 512
1653
+ dims: 7
1654
+ data_type: 1
1655
+ name: "model.10.body.2.weight"
1656
+ }
1657
+ initializer {
1658
+ dims: 512
1659
+ data_type: 1
1660
+ name: "model.11.body.0.bias"
1661
+ }
1662
+ initializer {
1663
+ dims: 512
1664
+ dims: 512
1665
+ dims: 7
1666
+ data_type: 1
1667
+ name: "model.11.body.0.weight"
1668
+ }
1669
+ initializer {
1670
+ dims: 512
1671
+ data_type: 1
1672
+ name: "model.11.body.2.bias"
1673
+ }
1674
+ initializer {
1675
+ dims: 512
1676
+ dims: 512
1677
+ dims: 7
1678
+ data_type: 1
1679
+ name: "model.11.body.2.weight"
1680
+ }
1681
+ initializer {
1682
+ dims: 512
1683
+ data_type: 1
1684
+ name: "model.12.body.0.bias"
1685
+ }
1686
+ initializer {
1687
+ dims: 512
1688
+ dims: 512
1689
+ dims: 7
1690
+ data_type: 1
1691
+ name: "model.12.body.0.weight"
1692
+ }
1693
+ initializer {
1694
+ dims: 512
1695
+ data_type: 1
1696
+ name: "model.12.body.2.bias"
1697
+ }
1698
+ initializer {
1699
+ dims: 512
1700
+ dims: 512
1701
+ dims: 7
1702
+ data_type: 1
1703
+ name: "model.12.body.2.weight"
1704
+ }
1705
+ initializer {
1706
+ dims: 512
1707
+ data_type: 1
1708
+ name: "model.13.body.0.bias"
1709
+ }
1710
+ initializer {
1711
+ dims: 512
1712
+ dims: 512
1713
+ dims: 7
1714
+ data_type: 1
1715
+ name: "model.13.body.0.weight"
1716
+ }
1717
+ initializer {
1718
+ dims: 512
1719
+ data_type: 1
1720
+ name: "model.13.body.2.bias"
1721
+ }
1722
+ initializer {
1723
+ dims: 512
1724
+ dims: 512
1725
+ dims: 7
1726
+ data_type: 1
1727
+ name: "model.13.body.2.weight"
1728
+ }
1729
+ initializer {
1730
+ dims: 512
1731
+ data_type: 1
1732
+ name: "model.14.body.0.bias"
1733
+ }
1734
+ initializer {
1735
+ dims: 512
1736
+ dims: 512
1737
+ dims: 7
1738
+ data_type: 1
1739
+ name: "model.14.body.0.weight"
1740
+ }
1741
+ initializer {
1742
+ dims: 512
1743
+ data_type: 1
1744
+ name: "model.14.body.2.bias"
1745
+ }
1746
+ initializer {
1747
+ dims: 512
1748
+ dims: 512
1749
+ dims: 7
1750
+ data_type: 1
1751
+ name: "model.14.body.2.weight"
1752
+ }
1753
+ initializer {
1754
+ dims: 512
1755
+ data_type: 1
1756
+ name: "model.15.body.0.bias"
1757
+ }
1758
+ initializer {
1759
+ dims: 512
1760
+ dims: 512
1761
+ dims: 7
1762
+ data_type: 1
1763
+ name: "model.15.body.0.weight"
1764
+ }
1765
+ initializer {
1766
+ dims: 512
1767
+ data_type: 1
1768
+ name: "model.15.body.2.bias"
1769
+ }
1770
+ initializer {
1771
+ dims: 512
1772
+ dims: 512
1773
+ dims: 7
1774
+ data_type: 1
1775
+ name: "model.15.body.2.weight"
1776
+ }
1777
+ initializer {
1778
+ dims: 2
1779
+ data_type: 1
1780
+ name: "model.16.bias"
1781
+ raw_data: "\215\230\224<-\260~\274"
1782
+ }
1783
+ initializer {
1784
+ dims: 2
1785
+ dims: 512
1786
+ dims: 1
1787
+ data_type: 1
1788
+ name: "model.16.weight"
1789
+ }
1790
+ initializer {
1791
+ dims: 512
1792
+ data_type: 1
1793
+ name: "model.2.body.0.bias"
1794
+ }
1795
+ initializer {
1796
+ dims: 512
1797
+ dims: 512
1798
+ dims: 7
1799
+ data_type: 1
1800
+ name: "model.2.body.0.weight"
1801
+ }
1802
+ initializer {
1803
+ dims: 512
1804
+ data_type: 1
1805
+ name: "model.2.body.2.bias"
1806
+ }
1807
+ initializer {
1808
+ dims: 512
1809
+ dims: 512
1810
+ dims: 7
1811
+ data_type: 1
1812
+ name: "model.2.body.2.weight"
1813
+ }
1814
+ initializer {
1815
+ dims: 512
1816
+ data_type: 1
1817
+ name: "model.3.body.0.bias"
1818
+ }
1819
+ initializer {
1820
+ dims: 512
1821
+ dims: 512
1822
+ dims: 7
1823
+ data_type: 1
1824
+ name: "model.3.body.0.weight"
1825
+ }
1826
+ initializer {
1827
+ dims: 512
1828
+ data_type: 1
1829
+ name: "model.3.body.2.bias"
1830
+ }
1831
+ initializer {
1832
+ dims: 512
1833
+ dims: 512
1834
+ dims: 7
1835
+ data_type: 1
1836
+ name: "model.3.body.2.weight"
1837
+ }
1838
+ initializer {
1839
+ dims: 512
1840
+ data_type: 1
1841
+ name: "model.4.body.0.bias"
1842
+ }
1843
+ initializer {
1844
+ dims: 512
1845
+ dims: 512
1846
+ dims: 7
1847
+ data_type: 1
1848
+ name: "model.4.body.0.weight"
1849
+ }
1850
+ initializer {
1851
+ dims: 512
1852
+ data_type: 1
1853
+ name: "model.4.body.2.bias"
1854
+ }
1855
+ initializer {
1856
+ dims: 512
1857
+ dims: 512
1858
+ dims: 7
1859
+ data_type: 1
1860
+ name: "model.4.body.2.weight"
1861
+ }
1862
+ initializer {
1863
+ dims: 512
1864
+ data_type: 1
1865
+ name: "model.5.body.0.bias"
1866
+ }
1867
+ initializer {
1868
+ dims: 512
1869
+ dims: 512
1870
+ dims: 7
1871
+ data_type: 1
1872
+ name: "model.5.body.0.weight"
1873
+ }
1874
+ initializer {
1875
+ dims: 512
1876
+ data_type: 1
1877
+ name: "model.5.body.2.bias"
1878
+ }
1879
+ initializer {
1880
+ dims: 512
1881
+ dims: 512
1882
+ dims: 7
1883
+ data_type: 1
1884
+ name: "model.5.body.2.weight"
1885
+ }
1886
+ initializer {
1887
+ dims: 512
1888
+ data_type: 1
1889
+ name: "model.6.body.0.bias"
1890
+ }
1891
+ initializer {
1892
+ dims: 512
1893
+ dims: 512
1894
+ dims: 7
1895
+ data_type: 1
1896
+ name: "model.6.body.0.weight"
1897
+ }
1898
+ initializer {
1899
+ dims: 512
1900
+ data_type: 1
1901
+ name: "model.6.body.2.bias"
1902
+ }
1903
+ initializer {
1904
+ dims: 512
1905
+ dims: 512
1906
+ dims: 7
1907
+ data_type: 1
1908
+ name: "model.6.body.2.weight"
1909
+ }
1910
+ initializer {
1911
+ dims: 512
1912
+ data_type: 1
1913
+ name: "model.7.body.0.bias"
1914
+ }
1915
+ initializer {
1916
+ dims: 512
1917
+ dims: 512
1918
+ dims: 7
1919
+ data_type: 1
1920
+ name: "model.7.body.0.weight"
1921
+ }
1922
+ initializer {
1923
+ dims: 512
1924
+ data_type: 1
1925
+ name: "model.7.body.2.bias"
1926
+ }
1927
+ initializer {
1928
+ dims: 512
1929
+ dims: 512
1930
+ dims: 7
1931
+ data_type: 1
1932
+ name: "model.7.body.2.weight"
1933
+ }
1934
+ initializer {
1935
+ dims: 512
1936
+ data_type: 1
1937
+ name: "model.8.body.0.bias"
1938
+ }
1939
+ initializer {
1940
+ dims: 512
1941
+ dims: 512
1942
+ dims: 7
1943
+ data_type: 1
1944
+ name: "model.8.body.0.weight"
1945
+ }
1946
+ initializer {
1947
+ dims: 512
1948
+ data_type: 1
1949
+ name: "model.8.body.2.bias"
1950
+ }
1951
+ initializer {
1952
+ dims: 512
1953
+ dims: 512
1954
+ dims: 7
1955
+ data_type: 1
1956
+ name: "model.8.body.2.weight"
1957
+ }
1958
+ initializer {
1959
+ dims: 512
1960
+ data_type: 1
1961
+ name: "model.9.body.0.bias"
1962
+ }
1963
+ initializer {
1964
+ dims: 512
1965
+ dims: 512
1966
+ dims: 7
1967
+ data_type: 1
1968
+ name: "model.9.body.0.weight"
1969
+ }
1970
+ initializer {
1971
+ dims: 512
1972
+ data_type: 1
1973
+ name: "model.9.body.2.bias"
1974
+ }
1975
+ initializer {
1976
+ dims: 512
1977
+ dims: 512
1978
+ dims: 7
1979
+ data_type: 1
1980
+ name: "model.9.body.2.weight"
1981
+ }
1982
+ input {
1983
+ name: "x"
1984
+ type {
1985
+ tensor_type {
1986
+ elem_type: 1
1987
+ shape {
1988
+ dim {
1989
+ dim_param: "batch_size"
1990
+ }
1991
+ dim {
1992
+ dim_value: 2
1993
+ }
1994
+ dim {
1995
+ dim_value: 8192
1996
+ }
1997
+ }
1998
+ }
1999
+ }
2000
+ }
2001
+ output {
2002
+ name: "y"
2003
+ type {
2004
+ tensor_type {
2005
+ elem_type: 1
2006
+ shape {
2007
+ dim {
2008
+ dim_param: "batch_size"
2009
+ }
2010
+ dim {
2011
+ dim_value: 2
2012
+ }
2013
+ dim {
2014
+ dim_value: 8192
2015
+ }
2016
+ }
2017
+ }
2018
+ }
2019
+ }
2020
+ }
2021
+ opset_import {
2022
+ version: 11
2023
+ }
models/ailia-models/resnetdo.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c3354e6b3c251c42d5305039d79b587e50cc905229236e93ae9301242a67610
3
+ size 220306739
models/ailia-models/resnetdo.onnx.prototxt ADDED
@@ -0,0 +1,2023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "model.0.weight"
10
+ input: "model.0.bias"
11
+ output: "65"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 7
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 3
32
+ ints: 3
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 1
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "65"
43
+ input: "model.1.body.0.weight"
44
+ input: "model.1.body.0.bias"
45
+ output: "66"
46
+ name: "Conv_1"
47
+ op_type: "Conv"
48
+ attribute {
49
+ name: "dilations"
50
+ ints: 1
51
+ type: INTS
52
+ }
53
+ attribute {
54
+ name: "group"
55
+ i: 1
56
+ type: INT
57
+ }
58
+ attribute {
59
+ name: "kernel_shape"
60
+ ints: 7
61
+ type: INTS
62
+ }
63
+ attribute {
64
+ name: "pads"
65
+ ints: 3
66
+ ints: 3
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "strides"
71
+ ints: 1
72
+ type: INTS
73
+ }
74
+ }
75
+ node {
76
+ input: "66"
77
+ output: "67"
78
+ name: "Relu_2"
79
+ op_type: "Relu"
80
+ }
81
+ node {
82
+ input: "67"
83
+ input: "model.1.body.3.weight"
84
+ input: "model.1.body.3.bias"
85
+ output: "68"
86
+ name: "Conv_3"
87
+ op_type: "Conv"
88
+ attribute {
89
+ name: "dilations"
90
+ ints: 1
91
+ type: INTS
92
+ }
93
+ attribute {
94
+ name: "group"
95
+ i: 1
96
+ type: INT
97
+ }
98
+ attribute {
99
+ name: "kernel_shape"
100
+ ints: 7
101
+ type: INTS
102
+ }
103
+ attribute {
104
+ name: "pads"
105
+ ints: 3
106
+ ints: 3
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "strides"
111
+ ints: 1
112
+ type: INTS
113
+ }
114
+ }
115
+ node {
116
+ output: "69"
117
+ name: "Constant_4"
118
+ op_type: "Constant"
119
+ attribute {
120
+ name: "value"
121
+ t {
122
+ data_type: 1
123
+ raw_data: "\315\314\314="
124
+ }
125
+ type: TENSOR
126
+ }
127
+ }
128
+ node {
129
+ input: "68"
130
+ input: "69"
131
+ output: "70"
132
+ name: "Mul_5"
133
+ op_type: "Mul"
134
+ }
135
+ node {
136
+ input: "70"
137
+ input: "65"
138
+ output: "71"
139
+ name: "Add_6"
140
+ op_type: "Add"
141
+ }
142
+ node {
143
+ input: "71"
144
+ input: "model.2.body.0.weight"
145
+ input: "model.2.body.0.bias"
146
+ output: "72"
147
+ name: "Conv_7"
148
+ op_type: "Conv"
149
+ attribute {
150
+ name: "dilations"
151
+ ints: 1
152
+ type: INTS
153
+ }
154
+ attribute {
155
+ name: "group"
156
+ i: 1
157
+ type: INT
158
+ }
159
+ attribute {
160
+ name: "kernel_shape"
161
+ ints: 7
162
+ type: INTS
163
+ }
164
+ attribute {
165
+ name: "pads"
166
+ ints: 3
167
+ ints: 3
168
+ type: INTS
169
+ }
170
+ attribute {
171
+ name: "strides"
172
+ ints: 1
173
+ type: INTS
174
+ }
175
+ }
176
+ node {
177
+ input: "72"
178
+ output: "73"
179
+ name: "Relu_8"
180
+ op_type: "Relu"
181
+ }
182
+ node {
183
+ input: "73"
184
+ input: "model.2.body.3.weight"
185
+ input: "model.2.body.3.bias"
186
+ output: "74"
187
+ name: "Conv_9"
188
+ op_type: "Conv"
189
+ attribute {
190
+ name: "dilations"
191
+ ints: 1
192
+ type: INTS
193
+ }
194
+ attribute {
195
+ name: "group"
196
+ i: 1
197
+ type: INT
198
+ }
199
+ attribute {
200
+ name: "kernel_shape"
201
+ ints: 7
202
+ type: INTS
203
+ }
204
+ attribute {
205
+ name: "pads"
206
+ ints: 3
207
+ ints: 3
208
+ type: INTS
209
+ }
210
+ attribute {
211
+ name: "strides"
212
+ ints: 1
213
+ type: INTS
214
+ }
215
+ }
216
+ node {
217
+ output: "75"
218
+ name: "Constant_10"
219
+ op_type: "Constant"
220
+ attribute {
221
+ name: "value"
222
+ t {
223
+ data_type: 1
224
+ raw_data: "\315\314\314="
225
+ }
226
+ type: TENSOR
227
+ }
228
+ }
229
+ node {
230
+ input: "74"
231
+ input: "75"
232
+ output: "76"
233
+ name: "Mul_11"
234
+ op_type: "Mul"
235
+ }
236
+ node {
237
+ input: "76"
238
+ input: "71"
239
+ output: "77"
240
+ name: "Add_12"
241
+ op_type: "Add"
242
+ }
243
+ node {
244
+ input: "77"
245
+ input: "model.3.body.0.weight"
246
+ input: "model.3.body.0.bias"
247
+ output: "78"
248
+ name: "Conv_13"
249
+ op_type: "Conv"
250
+ attribute {
251
+ name: "dilations"
252
+ ints: 1
253
+ type: INTS
254
+ }
255
+ attribute {
256
+ name: "group"
257
+ i: 1
258
+ type: INT
259
+ }
260
+ attribute {
261
+ name: "kernel_shape"
262
+ ints: 7
263
+ type: INTS
264
+ }
265
+ attribute {
266
+ name: "pads"
267
+ ints: 3
268
+ ints: 3
269
+ type: INTS
270
+ }
271
+ attribute {
272
+ name: "strides"
273
+ ints: 1
274
+ type: INTS
275
+ }
276
+ }
277
+ node {
278
+ input: "78"
279
+ output: "79"
280
+ name: "Relu_14"
281
+ op_type: "Relu"
282
+ }
283
+ node {
284
+ input: "79"
285
+ input: "model.3.body.3.weight"
286
+ input: "model.3.body.3.bias"
287
+ output: "80"
288
+ name: "Conv_15"
289
+ op_type: "Conv"
290
+ attribute {
291
+ name: "dilations"
292
+ ints: 1
293
+ type: INTS
294
+ }
295
+ attribute {
296
+ name: "group"
297
+ i: 1
298
+ type: INT
299
+ }
300
+ attribute {
301
+ name: "kernel_shape"
302
+ ints: 7
303
+ type: INTS
304
+ }
305
+ attribute {
306
+ name: "pads"
307
+ ints: 3
308
+ ints: 3
309
+ type: INTS
310
+ }
311
+ attribute {
312
+ name: "strides"
313
+ ints: 1
314
+ type: INTS
315
+ }
316
+ }
317
+ node {
318
+ output: "81"
319
+ name: "Constant_16"
320
+ op_type: "Constant"
321
+ attribute {
322
+ name: "value"
323
+ t {
324
+ data_type: 1
325
+ raw_data: "\315\314\314="
326
+ }
327
+ type: TENSOR
328
+ }
329
+ }
330
+ node {
331
+ input: "80"
332
+ input: "81"
333
+ output: "82"
334
+ name: "Mul_17"
335
+ op_type: "Mul"
336
+ }
337
+ node {
338
+ input: "82"
339
+ input: "77"
340
+ output: "83"
341
+ name: "Add_18"
342
+ op_type: "Add"
343
+ }
344
+ node {
345
+ input: "83"
346
+ input: "model.4.body.0.weight"
347
+ input: "model.4.body.0.bias"
348
+ output: "84"
349
+ name: "Conv_19"
350
+ op_type: "Conv"
351
+ attribute {
352
+ name: "dilations"
353
+ ints: 1
354
+ type: INTS
355
+ }
356
+ attribute {
357
+ name: "group"
358
+ i: 1
359
+ type: INT
360
+ }
361
+ attribute {
362
+ name: "kernel_shape"
363
+ ints: 7
364
+ type: INTS
365
+ }
366
+ attribute {
367
+ name: "pads"
368
+ ints: 3
369
+ ints: 3
370
+ type: INTS
371
+ }
372
+ attribute {
373
+ name: "strides"
374
+ ints: 1
375
+ type: INTS
376
+ }
377
+ }
378
+ node {
379
+ input: "84"
380
+ output: "85"
381
+ name: "Relu_20"
382
+ op_type: "Relu"
383
+ }
384
+ node {
385
+ input: "85"
386
+ input: "model.4.body.3.weight"
387
+ input: "model.4.body.3.bias"
388
+ output: "86"
389
+ name: "Conv_21"
390
+ op_type: "Conv"
391
+ attribute {
392
+ name: "dilations"
393
+ ints: 1
394
+ type: INTS
395
+ }
396
+ attribute {
397
+ name: "group"
398
+ i: 1
399
+ type: INT
400
+ }
401
+ attribute {
402
+ name: "kernel_shape"
403
+ ints: 7
404
+ type: INTS
405
+ }
406
+ attribute {
407
+ name: "pads"
408
+ ints: 3
409
+ ints: 3
410
+ type: INTS
411
+ }
412
+ attribute {
413
+ name: "strides"
414
+ ints: 1
415
+ type: INTS
416
+ }
417
+ }
418
+ node {
419
+ output: "87"
420
+ name: "Constant_22"
421
+ op_type: "Constant"
422
+ attribute {
423
+ name: "value"
424
+ t {
425
+ data_type: 1
426
+ raw_data: "\315\314\314="
427
+ }
428
+ type: TENSOR
429
+ }
430
+ }
431
+ node {
432
+ input: "86"
433
+ input: "87"
434
+ output: "88"
435
+ name: "Mul_23"
436
+ op_type: "Mul"
437
+ }
438
+ node {
439
+ input: "88"
440
+ input: "83"
441
+ output: "89"
442
+ name: "Add_24"
443
+ op_type: "Add"
444
+ }
445
+ node {
446
+ input: "89"
447
+ input: "model.5.body.0.weight"
448
+ input: "model.5.body.0.bias"
449
+ output: "90"
450
+ name: "Conv_25"
451
+ op_type: "Conv"
452
+ attribute {
453
+ name: "dilations"
454
+ ints: 1
455
+ type: INTS
456
+ }
457
+ attribute {
458
+ name: "group"
459
+ i: 1
460
+ type: INT
461
+ }
462
+ attribute {
463
+ name: "kernel_shape"
464
+ ints: 7
465
+ type: INTS
466
+ }
467
+ attribute {
468
+ name: "pads"
469
+ ints: 3
470
+ ints: 3
471
+ type: INTS
472
+ }
473
+ attribute {
474
+ name: "strides"
475
+ ints: 1
476
+ type: INTS
477
+ }
478
+ }
479
+ node {
480
+ input: "90"
481
+ output: "91"
482
+ name: "Relu_26"
483
+ op_type: "Relu"
484
+ }
485
+ node {
486
+ input: "91"
487
+ input: "model.5.body.3.weight"
488
+ input: "model.5.body.3.bias"
489
+ output: "92"
490
+ name: "Conv_27"
491
+ op_type: "Conv"
492
+ attribute {
493
+ name: "dilations"
494
+ ints: 1
495
+ type: INTS
496
+ }
497
+ attribute {
498
+ name: "group"
499
+ i: 1
500
+ type: INT
501
+ }
502
+ attribute {
503
+ name: "kernel_shape"
504
+ ints: 7
505
+ type: INTS
506
+ }
507
+ attribute {
508
+ name: "pads"
509
+ ints: 3
510
+ ints: 3
511
+ type: INTS
512
+ }
513
+ attribute {
514
+ name: "strides"
515
+ ints: 1
516
+ type: INTS
517
+ }
518
+ }
519
+ node {
520
+ output: "93"
521
+ name: "Constant_28"
522
+ op_type: "Constant"
523
+ attribute {
524
+ name: "value"
525
+ t {
526
+ data_type: 1
527
+ raw_data: "\315\314\314="
528
+ }
529
+ type: TENSOR
530
+ }
531
+ }
532
+ node {
533
+ input: "92"
534
+ input: "93"
535
+ output: "94"
536
+ name: "Mul_29"
537
+ op_type: "Mul"
538
+ }
539
+ node {
540
+ input: "94"
541
+ input: "89"
542
+ output: "95"
543
+ name: "Add_30"
544
+ op_type: "Add"
545
+ }
546
+ node {
547
+ input: "95"
548
+ input: "model.6.body.0.weight"
549
+ input: "model.6.body.0.bias"
550
+ output: "96"
551
+ name: "Conv_31"
552
+ op_type: "Conv"
553
+ attribute {
554
+ name: "dilations"
555
+ ints: 1
556
+ type: INTS
557
+ }
558
+ attribute {
559
+ name: "group"
560
+ i: 1
561
+ type: INT
562
+ }
563
+ attribute {
564
+ name: "kernel_shape"
565
+ ints: 7
566
+ type: INTS
567
+ }
568
+ attribute {
569
+ name: "pads"
570
+ ints: 3
571
+ ints: 3
572
+ type: INTS
573
+ }
574
+ attribute {
575
+ name: "strides"
576
+ ints: 1
577
+ type: INTS
578
+ }
579
+ }
580
+ node {
581
+ input: "96"
582
+ output: "97"
583
+ name: "Relu_32"
584
+ op_type: "Relu"
585
+ }
586
+ node {
587
+ input: "97"
588
+ input: "model.6.body.3.weight"
589
+ input: "model.6.body.3.bias"
590
+ output: "98"
591
+ name: "Conv_33"
592
+ op_type: "Conv"
593
+ attribute {
594
+ name: "dilations"
595
+ ints: 1
596
+ type: INTS
597
+ }
598
+ attribute {
599
+ name: "group"
600
+ i: 1
601
+ type: INT
602
+ }
603
+ attribute {
604
+ name: "kernel_shape"
605
+ ints: 7
606
+ type: INTS
607
+ }
608
+ attribute {
609
+ name: "pads"
610
+ ints: 3
611
+ ints: 3
612
+ type: INTS
613
+ }
614
+ attribute {
615
+ name: "strides"
616
+ ints: 1
617
+ type: INTS
618
+ }
619
+ }
620
+ node {
621
+ output: "99"
622
+ name: "Constant_34"
623
+ op_type: "Constant"
624
+ attribute {
625
+ name: "value"
626
+ t {
627
+ data_type: 1
628
+ raw_data: "\315\314\314="
629
+ }
630
+ type: TENSOR
631
+ }
632
+ }
633
+ node {
634
+ input: "98"
635
+ input: "99"
636
+ output: "100"
637
+ name: "Mul_35"
638
+ op_type: "Mul"
639
+ }
640
+ node {
641
+ input: "100"
642
+ input: "95"
643
+ output: "101"
644
+ name: "Add_36"
645
+ op_type: "Add"
646
+ }
647
+ node {
648
+ input: "101"
649
+ input: "model.7.body.0.weight"
650
+ input: "model.7.body.0.bias"
651
+ output: "102"
652
+ name: "Conv_37"
653
+ op_type: "Conv"
654
+ attribute {
655
+ name: "dilations"
656
+ ints: 1
657
+ type: INTS
658
+ }
659
+ attribute {
660
+ name: "group"
661
+ i: 1
662
+ type: INT
663
+ }
664
+ attribute {
665
+ name: "kernel_shape"
666
+ ints: 7
667
+ type: INTS
668
+ }
669
+ attribute {
670
+ name: "pads"
671
+ ints: 3
672
+ ints: 3
673
+ type: INTS
674
+ }
675
+ attribute {
676
+ name: "strides"
677
+ ints: 1
678
+ type: INTS
679
+ }
680
+ }
681
+ node {
682
+ input: "102"
683
+ output: "103"
684
+ name: "Relu_38"
685
+ op_type: "Relu"
686
+ }
687
+ node {
688
+ input: "103"
689
+ input: "model.7.body.3.weight"
690
+ input: "model.7.body.3.bias"
691
+ output: "104"
692
+ name: "Conv_39"
693
+ op_type: "Conv"
694
+ attribute {
695
+ name: "dilations"
696
+ ints: 1
697
+ type: INTS
698
+ }
699
+ attribute {
700
+ name: "group"
701
+ i: 1
702
+ type: INT
703
+ }
704
+ attribute {
705
+ name: "kernel_shape"
706
+ ints: 7
707
+ type: INTS
708
+ }
709
+ attribute {
710
+ name: "pads"
711
+ ints: 3
712
+ ints: 3
713
+ type: INTS
714
+ }
715
+ attribute {
716
+ name: "strides"
717
+ ints: 1
718
+ type: INTS
719
+ }
720
+ }
721
+ node {
722
+ output: "105"
723
+ name: "Constant_40"
724
+ op_type: "Constant"
725
+ attribute {
726
+ name: "value"
727
+ t {
728
+ data_type: 1
729
+ raw_data: "\315\314\314="
730
+ }
731
+ type: TENSOR
732
+ }
733
+ }
734
+ node {
735
+ input: "104"
736
+ input: "105"
737
+ output: "106"
738
+ name: "Mul_41"
739
+ op_type: "Mul"
740
+ }
741
+ node {
742
+ input: "106"
743
+ input: "101"
744
+ output: "107"
745
+ name: "Add_42"
746
+ op_type: "Add"
747
+ }
748
+ node {
749
+ input: "107"
750
+ input: "model.8.body.0.weight"
751
+ input: "model.8.body.0.bias"
752
+ output: "108"
753
+ name: "Conv_43"
754
+ op_type: "Conv"
755
+ attribute {
756
+ name: "dilations"
757
+ ints: 1
758
+ type: INTS
759
+ }
760
+ attribute {
761
+ name: "group"
762
+ i: 1
763
+ type: INT
764
+ }
765
+ attribute {
766
+ name: "kernel_shape"
767
+ ints: 7
768
+ type: INTS
769
+ }
770
+ attribute {
771
+ name: "pads"
772
+ ints: 3
773
+ ints: 3
774
+ type: INTS
775
+ }
776
+ attribute {
777
+ name: "strides"
778
+ ints: 1
779
+ type: INTS
780
+ }
781
+ }
782
+ node {
783
+ input: "108"
784
+ output: "109"
785
+ name: "Relu_44"
786
+ op_type: "Relu"
787
+ }
788
+ node {
789
+ input: "109"
790
+ input: "model.8.body.3.weight"
791
+ input: "model.8.body.3.bias"
792
+ output: "110"
793
+ name: "Conv_45"
794
+ op_type: "Conv"
795
+ attribute {
796
+ name: "dilations"
797
+ ints: 1
798
+ type: INTS
799
+ }
800
+ attribute {
801
+ name: "group"
802
+ i: 1
803
+ type: INT
804
+ }
805
+ attribute {
806
+ name: "kernel_shape"
807
+ ints: 7
808
+ type: INTS
809
+ }
810
+ attribute {
811
+ name: "pads"
812
+ ints: 3
813
+ ints: 3
814
+ type: INTS
815
+ }
816
+ attribute {
817
+ name: "strides"
818
+ ints: 1
819
+ type: INTS
820
+ }
821
+ }
822
+ node {
823
+ output: "111"
824
+ name: "Constant_46"
825
+ op_type: "Constant"
826
+ attribute {
827
+ name: "value"
828
+ t {
829
+ data_type: 1
830
+ raw_data: "\315\314\314="
831
+ }
832
+ type: TENSOR
833
+ }
834
+ }
835
+ node {
836
+ input: "110"
837
+ input: "111"
838
+ output: "112"
839
+ name: "Mul_47"
840
+ op_type: "Mul"
841
+ }
842
+ node {
843
+ input: "112"
844
+ input: "107"
845
+ output: "113"
846
+ name: "Add_48"
847
+ op_type: "Add"
848
+ }
849
+ node {
850
+ input: "113"
851
+ input: "model.9.body.0.weight"
852
+ input: "model.9.body.0.bias"
853
+ output: "114"
854
+ name: "Conv_49"
855
+ op_type: "Conv"
856
+ attribute {
857
+ name: "dilations"
858
+ ints: 1
859
+ type: INTS
860
+ }
861
+ attribute {
862
+ name: "group"
863
+ i: 1
864
+ type: INT
865
+ }
866
+ attribute {
867
+ name: "kernel_shape"
868
+ ints: 7
869
+ type: INTS
870
+ }
871
+ attribute {
872
+ name: "pads"
873
+ ints: 3
874
+ ints: 3
875
+ type: INTS
876
+ }
877
+ attribute {
878
+ name: "strides"
879
+ ints: 1
880
+ type: INTS
881
+ }
882
+ }
883
+ node {
884
+ input: "114"
885
+ output: "115"
886
+ name: "Relu_50"
887
+ op_type: "Relu"
888
+ }
889
+ node {
890
+ input: "115"
891
+ input: "model.9.body.3.weight"
892
+ input: "model.9.body.3.bias"
893
+ output: "116"
894
+ name: "Conv_51"
895
+ op_type: "Conv"
896
+ attribute {
897
+ name: "dilations"
898
+ ints: 1
899
+ type: INTS
900
+ }
901
+ attribute {
902
+ name: "group"
903
+ i: 1
904
+ type: INT
905
+ }
906
+ attribute {
907
+ name: "kernel_shape"
908
+ ints: 7
909
+ type: INTS
910
+ }
911
+ attribute {
912
+ name: "pads"
913
+ ints: 3
914
+ ints: 3
915
+ type: INTS
916
+ }
917
+ attribute {
918
+ name: "strides"
919
+ ints: 1
920
+ type: INTS
921
+ }
922
+ }
923
+ node {
924
+ output: "117"
925
+ name: "Constant_52"
926
+ op_type: "Constant"
927
+ attribute {
928
+ name: "value"
929
+ t {
930
+ data_type: 1
931
+ raw_data: "\315\314\314="
932
+ }
933
+ type: TENSOR
934
+ }
935
+ }
936
+ node {
937
+ input: "116"
938
+ input: "117"
939
+ output: "118"
940
+ name: "Mul_53"
941
+ op_type: "Mul"
942
+ }
943
+ node {
944
+ input: "118"
945
+ input: "113"
946
+ output: "119"
947
+ name: "Add_54"
948
+ op_type: "Add"
949
+ }
950
+ node {
951
+ input: "119"
952
+ input: "model.10.body.0.weight"
953
+ input: "model.10.body.0.bias"
954
+ output: "120"
955
+ name: "Conv_55"
956
+ op_type: "Conv"
957
+ attribute {
958
+ name: "dilations"
959
+ ints: 1
960
+ type: INTS
961
+ }
962
+ attribute {
963
+ name: "group"
964
+ i: 1
965
+ type: INT
966
+ }
967
+ attribute {
968
+ name: "kernel_shape"
969
+ ints: 7
970
+ type: INTS
971
+ }
972
+ attribute {
973
+ name: "pads"
974
+ ints: 3
975
+ ints: 3
976
+ type: INTS
977
+ }
978
+ attribute {
979
+ name: "strides"
980
+ ints: 1
981
+ type: INTS
982
+ }
983
+ }
984
+ node {
985
+ input: "120"
986
+ output: "121"
987
+ name: "Relu_56"
988
+ op_type: "Relu"
989
+ }
990
+ node {
991
+ input: "121"
992
+ input: "model.10.body.3.weight"
993
+ input: "model.10.body.3.bias"
994
+ output: "122"
995
+ name: "Conv_57"
996
+ op_type: "Conv"
997
+ attribute {
998
+ name: "dilations"
999
+ ints: 1
1000
+ type: INTS
1001
+ }
1002
+ attribute {
1003
+ name: "group"
1004
+ i: 1
1005
+ type: INT
1006
+ }
1007
+ attribute {
1008
+ name: "kernel_shape"
1009
+ ints: 7
1010
+ type: INTS
1011
+ }
1012
+ attribute {
1013
+ name: "pads"
1014
+ ints: 3
1015
+ ints: 3
1016
+ type: INTS
1017
+ }
1018
+ attribute {
1019
+ name: "strides"
1020
+ ints: 1
1021
+ type: INTS
1022
+ }
1023
+ }
1024
+ node {
1025
+ output: "123"
1026
+ name: "Constant_58"
1027
+ op_type: "Constant"
1028
+ attribute {
1029
+ name: "value"
1030
+ t {
1031
+ data_type: 1
1032
+ raw_data: "\315\314\314="
1033
+ }
1034
+ type: TENSOR
1035
+ }
1036
+ }
1037
+ node {
1038
+ input: "122"
1039
+ input: "123"
1040
+ output: "124"
1041
+ name: "Mul_59"
1042
+ op_type: "Mul"
1043
+ }
1044
+ node {
1045
+ input: "124"
1046
+ input: "119"
1047
+ output: "125"
1048
+ name: "Add_60"
1049
+ op_type: "Add"
1050
+ }
1051
+ node {
1052
+ input: "125"
1053
+ input: "model.11.body.0.weight"
1054
+ input: "model.11.body.0.bias"
1055
+ output: "126"
1056
+ name: "Conv_61"
1057
+ op_type: "Conv"
1058
+ attribute {
1059
+ name: "dilations"
1060
+ ints: 1
1061
+ type: INTS
1062
+ }
1063
+ attribute {
1064
+ name: "group"
1065
+ i: 1
1066
+ type: INT
1067
+ }
1068
+ attribute {
1069
+ name: "kernel_shape"
1070
+ ints: 7
1071
+ type: INTS
1072
+ }
1073
+ attribute {
1074
+ name: "pads"
1075
+ ints: 3
1076
+ ints: 3
1077
+ type: INTS
1078
+ }
1079
+ attribute {
1080
+ name: "strides"
1081
+ ints: 1
1082
+ type: INTS
1083
+ }
1084
+ }
1085
+ node {
1086
+ input: "126"
1087
+ output: "127"
1088
+ name: "Relu_62"
1089
+ op_type: "Relu"
1090
+ }
1091
+ node {
1092
+ input: "127"
1093
+ input: "model.11.body.3.weight"
1094
+ input: "model.11.body.3.bias"
1095
+ output: "128"
1096
+ name: "Conv_63"
1097
+ op_type: "Conv"
1098
+ attribute {
1099
+ name: "dilations"
1100
+ ints: 1
1101
+ type: INTS
1102
+ }
1103
+ attribute {
1104
+ name: "group"
1105
+ i: 1
1106
+ type: INT
1107
+ }
1108
+ attribute {
1109
+ name: "kernel_shape"
1110
+ ints: 7
1111
+ type: INTS
1112
+ }
1113
+ attribute {
1114
+ name: "pads"
1115
+ ints: 3
1116
+ ints: 3
1117
+ type: INTS
1118
+ }
1119
+ attribute {
1120
+ name: "strides"
1121
+ ints: 1
1122
+ type: INTS
1123
+ }
1124
+ }
1125
+ node {
1126
+ output: "129"
1127
+ name: "Constant_64"
1128
+ op_type: "Constant"
1129
+ attribute {
1130
+ name: "value"
1131
+ t {
1132
+ data_type: 1
1133
+ raw_data: "\315\314\314="
1134
+ }
1135
+ type: TENSOR
1136
+ }
1137
+ }
1138
+ node {
1139
+ input: "128"
1140
+ input: "129"
1141
+ output: "130"
1142
+ name: "Mul_65"
1143
+ op_type: "Mul"
1144
+ }
1145
+ node {
1146
+ input: "130"
1147
+ input: "125"
1148
+ output: "131"
1149
+ name: "Add_66"
1150
+ op_type: "Add"
1151
+ }
1152
+ node {
1153
+ input: "131"
1154
+ input: "model.12.body.0.weight"
1155
+ input: "model.12.body.0.bias"
1156
+ output: "132"
1157
+ name: "Conv_67"
1158
+ op_type: "Conv"
1159
+ attribute {
1160
+ name: "dilations"
1161
+ ints: 1
1162
+ type: INTS
1163
+ }
1164
+ attribute {
1165
+ name: "group"
1166
+ i: 1
1167
+ type: INT
1168
+ }
1169
+ attribute {
1170
+ name: "kernel_shape"
1171
+ ints: 7
1172
+ type: INTS
1173
+ }
1174
+ attribute {
1175
+ name: "pads"
1176
+ ints: 3
1177
+ ints: 3
1178
+ type: INTS
1179
+ }
1180
+ attribute {
1181
+ name: "strides"
1182
+ ints: 1
1183
+ type: INTS
1184
+ }
1185
+ }
1186
+ node {
1187
+ input: "132"
1188
+ output: "133"
1189
+ name: "Relu_68"
1190
+ op_type: "Relu"
1191
+ }
1192
+ node {
1193
+ input: "133"
1194
+ input: "model.12.body.3.weight"
1195
+ input: "model.12.body.3.bias"
1196
+ output: "134"
1197
+ name: "Conv_69"
1198
+ op_type: "Conv"
1199
+ attribute {
1200
+ name: "dilations"
1201
+ ints: 1
1202
+ type: INTS
1203
+ }
1204
+ attribute {
1205
+ name: "group"
1206
+ i: 1
1207
+ type: INT
1208
+ }
1209
+ attribute {
1210
+ name: "kernel_shape"
1211
+ ints: 7
1212
+ type: INTS
1213
+ }
1214
+ attribute {
1215
+ name: "pads"
1216
+ ints: 3
1217
+ ints: 3
1218
+ type: INTS
1219
+ }
1220
+ attribute {
1221
+ name: "strides"
1222
+ ints: 1
1223
+ type: INTS
1224
+ }
1225
+ }
1226
+ node {
1227
+ output: "135"
1228
+ name: "Constant_70"
1229
+ op_type: "Constant"
1230
+ attribute {
1231
+ name: "value"
1232
+ t {
1233
+ data_type: 1
1234
+ raw_data: "\315\314\314="
1235
+ }
1236
+ type: TENSOR
1237
+ }
1238
+ }
1239
+ node {
1240
+ input: "134"
1241
+ input: "135"
1242
+ output: "136"
1243
+ name: "Mul_71"
1244
+ op_type: "Mul"
1245
+ }
1246
+ node {
1247
+ input: "136"
1248
+ input: "131"
1249
+ output: "137"
1250
+ name: "Add_72"
1251
+ op_type: "Add"
1252
+ }
1253
+ node {
1254
+ input: "137"
1255
+ input: "model.13.body.0.weight"
1256
+ input: "model.13.body.0.bias"
1257
+ output: "138"
1258
+ name: "Conv_73"
1259
+ op_type: "Conv"
1260
+ attribute {
1261
+ name: "dilations"
1262
+ ints: 1
1263
+ type: INTS
1264
+ }
1265
+ attribute {
1266
+ name: "group"
1267
+ i: 1
1268
+ type: INT
1269
+ }
1270
+ attribute {
1271
+ name: "kernel_shape"
1272
+ ints: 7
1273
+ type: INTS
1274
+ }
1275
+ attribute {
1276
+ name: "pads"
1277
+ ints: 3
1278
+ ints: 3
1279
+ type: INTS
1280
+ }
1281
+ attribute {
1282
+ name: "strides"
1283
+ ints: 1
1284
+ type: INTS
1285
+ }
1286
+ }
1287
+ node {
1288
+ input: "138"
1289
+ output: "139"
1290
+ name: "Relu_74"
1291
+ op_type: "Relu"
1292
+ }
1293
+ node {
1294
+ input: "139"
1295
+ input: "model.13.body.3.weight"
1296
+ input: "model.13.body.3.bias"
1297
+ output: "140"
1298
+ name: "Conv_75"
1299
+ op_type: "Conv"
1300
+ attribute {
1301
+ name: "dilations"
1302
+ ints: 1
1303
+ type: INTS
1304
+ }
1305
+ attribute {
1306
+ name: "group"
1307
+ i: 1
1308
+ type: INT
1309
+ }
1310
+ attribute {
1311
+ name: "kernel_shape"
1312
+ ints: 7
1313
+ type: INTS
1314
+ }
1315
+ attribute {
1316
+ name: "pads"
1317
+ ints: 3
1318
+ ints: 3
1319
+ type: INTS
1320
+ }
1321
+ attribute {
1322
+ name: "strides"
1323
+ ints: 1
1324
+ type: INTS
1325
+ }
1326
+ }
1327
+ node {
1328
+ output: "141"
1329
+ name: "Constant_76"
1330
+ op_type: "Constant"
1331
+ attribute {
1332
+ name: "value"
1333
+ t {
1334
+ data_type: 1
1335
+ raw_data: "\315\314\314="
1336
+ }
1337
+ type: TENSOR
1338
+ }
1339
+ }
1340
+ node {
1341
+ input: "140"
1342
+ input: "141"
1343
+ output: "142"
1344
+ name: "Mul_77"
1345
+ op_type: "Mul"
1346
+ }
1347
+ node {
1348
+ input: "142"
1349
+ input: "137"
1350
+ output: "143"
1351
+ name: "Add_78"
1352
+ op_type: "Add"
1353
+ }
1354
+ node {
1355
+ input: "143"
1356
+ input: "model.14.body.0.weight"
1357
+ input: "model.14.body.0.bias"
1358
+ output: "144"
1359
+ name: "Conv_79"
1360
+ op_type: "Conv"
1361
+ attribute {
1362
+ name: "dilations"
1363
+ ints: 1
1364
+ type: INTS
1365
+ }
1366
+ attribute {
1367
+ name: "group"
1368
+ i: 1
1369
+ type: INT
1370
+ }
1371
+ attribute {
1372
+ name: "kernel_shape"
1373
+ ints: 7
1374
+ type: INTS
1375
+ }
1376
+ attribute {
1377
+ name: "pads"
1378
+ ints: 3
1379
+ ints: 3
1380
+ type: INTS
1381
+ }
1382
+ attribute {
1383
+ name: "strides"
1384
+ ints: 1
1385
+ type: INTS
1386
+ }
1387
+ }
1388
+ node {
1389
+ input: "144"
1390
+ output: "145"
1391
+ name: "Relu_80"
1392
+ op_type: "Relu"
1393
+ }
1394
+ node {
1395
+ input: "145"
1396
+ input: "model.14.body.3.weight"
1397
+ input: "model.14.body.3.bias"
1398
+ output: "146"
1399
+ name: "Conv_81"
1400
+ op_type: "Conv"
1401
+ attribute {
1402
+ name: "dilations"
1403
+ ints: 1
1404
+ type: INTS
1405
+ }
1406
+ attribute {
1407
+ name: "group"
1408
+ i: 1
1409
+ type: INT
1410
+ }
1411
+ attribute {
1412
+ name: "kernel_shape"
1413
+ ints: 7
1414
+ type: INTS
1415
+ }
1416
+ attribute {
1417
+ name: "pads"
1418
+ ints: 3
1419
+ ints: 3
1420
+ type: INTS
1421
+ }
1422
+ attribute {
1423
+ name: "strides"
1424
+ ints: 1
1425
+ type: INTS
1426
+ }
1427
+ }
1428
+ node {
1429
+ output: "147"
1430
+ name: "Constant_82"
1431
+ op_type: "Constant"
1432
+ attribute {
1433
+ name: "value"
1434
+ t {
1435
+ data_type: 1
1436
+ raw_data: "\315\314\314="
1437
+ }
1438
+ type: TENSOR
1439
+ }
1440
+ }
1441
+ node {
1442
+ input: "146"
1443
+ input: "147"
1444
+ output: "148"
1445
+ name: "Mul_83"
1446
+ op_type: "Mul"
1447
+ }
1448
+ node {
1449
+ input: "148"
1450
+ input: "143"
1451
+ output: "149"
1452
+ name: "Add_84"
1453
+ op_type: "Add"
1454
+ }
1455
+ node {
1456
+ input: "149"
1457
+ input: "model.15.body.0.weight"
1458
+ input: "model.15.body.0.bias"
1459
+ output: "150"
1460
+ name: "Conv_85"
1461
+ op_type: "Conv"
1462
+ attribute {
1463
+ name: "dilations"
1464
+ ints: 1
1465
+ type: INTS
1466
+ }
1467
+ attribute {
1468
+ name: "group"
1469
+ i: 1
1470
+ type: INT
1471
+ }
1472
+ attribute {
1473
+ name: "kernel_shape"
1474
+ ints: 7
1475
+ type: INTS
1476
+ }
1477
+ attribute {
1478
+ name: "pads"
1479
+ ints: 3
1480
+ ints: 3
1481
+ type: INTS
1482
+ }
1483
+ attribute {
1484
+ name: "strides"
1485
+ ints: 1
1486
+ type: INTS
1487
+ }
1488
+ }
1489
+ node {
1490
+ input: "150"
1491
+ output: "151"
1492
+ name: "Relu_86"
1493
+ op_type: "Relu"
1494
+ }
1495
+ node {
1496
+ input: "151"
1497
+ input: "model.15.body.3.weight"
1498
+ input: "model.15.body.3.bias"
1499
+ output: "152"
1500
+ name: "Conv_87"
1501
+ op_type: "Conv"
1502
+ attribute {
1503
+ name: "dilations"
1504
+ ints: 1
1505
+ type: INTS
1506
+ }
1507
+ attribute {
1508
+ name: "group"
1509
+ i: 1
1510
+ type: INT
1511
+ }
1512
+ attribute {
1513
+ name: "kernel_shape"
1514
+ ints: 7
1515
+ type: INTS
1516
+ }
1517
+ attribute {
1518
+ name: "pads"
1519
+ ints: 3
1520
+ ints: 3
1521
+ type: INTS
1522
+ }
1523
+ attribute {
1524
+ name: "strides"
1525
+ ints: 1
1526
+ type: INTS
1527
+ }
1528
+ }
1529
+ node {
1530
+ output: "153"
1531
+ name: "Constant_88"
1532
+ op_type: "Constant"
1533
+ attribute {
1534
+ name: "value"
1535
+ t {
1536
+ data_type: 1
1537
+ raw_data: "\315\314\314="
1538
+ }
1539
+ type: TENSOR
1540
+ }
1541
+ }
1542
+ node {
1543
+ input: "152"
1544
+ input: "153"
1545
+ output: "154"
1546
+ name: "Mul_89"
1547
+ op_type: "Mul"
1548
+ }
1549
+ node {
1550
+ input: "154"
1551
+ input: "149"
1552
+ output: "155"
1553
+ name: "Add_90"
1554
+ op_type: "Add"
1555
+ }
1556
+ node {
1557
+ input: "155"
1558
+ input: "model.16.weight"
1559
+ input: "model.16.bias"
1560
+ output: "156"
1561
+ name: "Conv_91"
1562
+ op_type: "Conv"
1563
+ attribute {
1564
+ name: "dilations"
1565
+ ints: 1
1566
+ type: INTS
1567
+ }
1568
+ attribute {
1569
+ name: "group"
1570
+ i: 1
1571
+ type: INT
1572
+ }
1573
+ attribute {
1574
+ name: "kernel_shape"
1575
+ ints: 1
1576
+ type: INTS
1577
+ }
1578
+ attribute {
1579
+ name: "pads"
1580
+ ints: 0
1581
+ ints: 0
1582
+ type: INTS
1583
+ }
1584
+ attribute {
1585
+ name: "strides"
1586
+ ints: 1
1587
+ type: INTS
1588
+ }
1589
+ }
1590
+ node {
1591
+ input: "156"
1592
+ input: "x"
1593
+ output: "y"
1594
+ name: "Add_92"
1595
+ op_type: "Add"
1596
+ }
1597
+ initializer {
1598
+ dims: 512
1599
+ data_type: 1
1600
+ name: "model.0.bias"
1601
+ }
1602
+ initializer {
1603
+ dims: 512
1604
+ dims: 2
1605
+ dims: 7
1606
+ data_type: 1
1607
+ name: "model.0.weight"
1608
+ }
1609
+ initializer {
1610
+ dims: 512
1611
+ data_type: 1
1612
+ name: "model.1.body.0.bias"
1613
+ }
1614
+ initializer {
1615
+ dims: 512
1616
+ dims: 512
1617
+ dims: 7
1618
+ data_type: 1
1619
+ name: "model.1.body.0.weight"
1620
+ }
1621
+ initializer {
1622
+ dims: 512
1623
+ data_type: 1
1624
+ name: "model.1.body.3.bias"
1625
+ }
1626
+ initializer {
1627
+ dims: 512
1628
+ dims: 512
1629
+ dims: 7
1630
+ data_type: 1
1631
+ name: "model.1.body.3.weight"
1632
+ }
1633
+ initializer {
1634
+ dims: 512
1635
+ data_type: 1
1636
+ name: "model.10.body.0.bias"
1637
+ }
1638
+ initializer {
1639
+ dims: 512
1640
+ dims: 512
1641
+ dims: 7
1642
+ data_type: 1
1643
+ name: "model.10.body.0.weight"
1644
+ }
1645
+ initializer {
1646
+ dims: 512
1647
+ data_type: 1
1648
+ name: "model.10.body.3.bias"
1649
+ }
1650
+ initializer {
1651
+ dims: 512
1652
+ dims: 512
1653
+ dims: 7
1654
+ data_type: 1
1655
+ name: "model.10.body.3.weight"
1656
+ }
1657
+ initializer {
1658
+ dims: 512
1659
+ data_type: 1
1660
+ name: "model.11.body.0.bias"
1661
+ }
1662
+ initializer {
1663
+ dims: 512
1664
+ dims: 512
1665
+ dims: 7
1666
+ data_type: 1
1667
+ name: "model.11.body.0.weight"
1668
+ }
1669
+ initializer {
1670
+ dims: 512
1671
+ data_type: 1
1672
+ name: "model.11.body.3.bias"
1673
+ }
1674
+ initializer {
1675
+ dims: 512
1676
+ dims: 512
1677
+ dims: 7
1678
+ data_type: 1
1679
+ name: "model.11.body.3.weight"
1680
+ }
1681
+ initializer {
1682
+ dims: 512
1683
+ data_type: 1
1684
+ name: "model.12.body.0.bias"
1685
+ }
1686
+ initializer {
1687
+ dims: 512
1688
+ dims: 512
1689
+ dims: 7
1690
+ data_type: 1
1691
+ name: "model.12.body.0.weight"
1692
+ }
1693
+ initializer {
1694
+ dims: 512
1695
+ data_type: 1
1696
+ name: "model.12.body.3.bias"
1697
+ }
1698
+ initializer {
1699
+ dims: 512
1700
+ dims: 512
1701
+ dims: 7
1702
+ data_type: 1
1703
+ name: "model.12.body.3.weight"
1704
+ }
1705
+ initializer {
1706
+ dims: 512
1707
+ data_type: 1
1708
+ name: "model.13.body.0.bias"
1709
+ }
1710
+ initializer {
1711
+ dims: 512
1712
+ dims: 512
1713
+ dims: 7
1714
+ data_type: 1
1715
+ name: "model.13.body.0.weight"
1716
+ }
1717
+ initializer {
1718
+ dims: 512
1719
+ data_type: 1
1720
+ name: "model.13.body.3.bias"
1721
+ }
1722
+ initializer {
1723
+ dims: 512
1724
+ dims: 512
1725
+ dims: 7
1726
+ data_type: 1
1727
+ name: "model.13.body.3.weight"
1728
+ }
1729
+ initializer {
1730
+ dims: 512
1731
+ data_type: 1
1732
+ name: "model.14.body.0.bias"
1733
+ }
1734
+ initializer {
1735
+ dims: 512
1736
+ dims: 512
1737
+ dims: 7
1738
+ data_type: 1
1739
+ name: "model.14.body.0.weight"
1740
+ }
1741
+ initializer {
1742
+ dims: 512
1743
+ data_type: 1
1744
+ name: "model.14.body.3.bias"
1745
+ }
1746
+ initializer {
1747
+ dims: 512
1748
+ dims: 512
1749
+ dims: 7
1750
+ data_type: 1
1751
+ name: "model.14.body.3.weight"
1752
+ }
1753
+ initializer {
1754
+ dims: 512
1755
+ data_type: 1
1756
+ name: "model.15.body.0.bias"
1757
+ }
1758
+ initializer {
1759
+ dims: 512
1760
+ dims: 512
1761
+ dims: 7
1762
+ data_type: 1
1763
+ name: "model.15.body.0.weight"
1764
+ }
1765
+ initializer {
1766
+ dims: 512
1767
+ data_type: 1
1768
+ name: "model.15.body.3.bias"
1769
+ }
1770
+ initializer {
1771
+ dims: 512
1772
+ dims: 512
1773
+ dims: 7
1774
+ data_type: 1
1775
+ name: "model.15.body.3.weight"
1776
+ }
1777
+ initializer {
1778
+ dims: 2
1779
+ data_type: 1
1780
+ name: "model.16.bias"
1781
+ raw_data: "\t\226\031\275\337\257\321<"
1782
+ }
1783
+ initializer {
1784
+ dims: 2
1785
+ dims: 512
1786
+ dims: 1
1787
+ data_type: 1
1788
+ name: "model.16.weight"
1789
+ }
1790
+ initializer {
1791
+ dims: 512
1792
+ data_type: 1
1793
+ name: "model.2.body.0.bias"
1794
+ }
1795
+ initializer {
1796
+ dims: 512
1797
+ dims: 512
1798
+ dims: 7
1799
+ data_type: 1
1800
+ name: "model.2.body.0.weight"
1801
+ }
1802
+ initializer {
1803
+ dims: 512
1804
+ data_type: 1
1805
+ name: "model.2.body.3.bias"
1806
+ }
1807
+ initializer {
1808
+ dims: 512
1809
+ dims: 512
1810
+ dims: 7
1811
+ data_type: 1
1812
+ name: "model.2.body.3.weight"
1813
+ }
1814
+ initializer {
1815
+ dims: 512
1816
+ data_type: 1
1817
+ name: "model.3.body.0.bias"
1818
+ }
1819
+ initializer {
1820
+ dims: 512
1821
+ dims: 512
1822
+ dims: 7
1823
+ data_type: 1
1824
+ name: "model.3.body.0.weight"
1825
+ }
1826
+ initializer {
1827
+ dims: 512
1828
+ data_type: 1
1829
+ name: "model.3.body.3.bias"
1830
+ }
1831
+ initializer {
1832
+ dims: 512
1833
+ dims: 512
1834
+ dims: 7
1835
+ data_type: 1
1836
+ name: "model.3.body.3.weight"
1837
+ }
1838
+ initializer {
1839
+ dims: 512
1840
+ data_type: 1
1841
+ name: "model.4.body.0.bias"
1842
+ }
1843
+ initializer {
1844
+ dims: 512
1845
+ dims: 512
1846
+ dims: 7
1847
+ data_type: 1
1848
+ name: "model.4.body.0.weight"
1849
+ }
1850
+ initializer {
1851
+ dims: 512
1852
+ data_type: 1
1853
+ name: "model.4.body.3.bias"
1854
+ }
1855
+ initializer {
1856
+ dims: 512
1857
+ dims: 512
1858
+ dims: 7
1859
+ data_type: 1
1860
+ name: "model.4.body.3.weight"
1861
+ }
1862
+ initializer {
1863
+ dims: 512
1864
+ data_type: 1
1865
+ name: "model.5.body.0.bias"
1866
+ }
1867
+ initializer {
1868
+ dims: 512
1869
+ dims: 512
1870
+ dims: 7
1871
+ data_type: 1
1872
+ name: "model.5.body.0.weight"
1873
+ }
1874
+ initializer {
1875
+ dims: 512
1876
+ data_type: 1
1877
+ name: "model.5.body.3.bias"
1878
+ }
1879
+ initializer {
1880
+ dims: 512
1881
+ dims: 512
1882
+ dims: 7
1883
+ data_type: 1
1884
+ name: "model.5.body.3.weight"
1885
+ }
1886
+ initializer {
1887
+ dims: 512
1888
+ data_type: 1
1889
+ name: "model.6.body.0.bias"
1890
+ }
1891
+ initializer {
1892
+ dims: 512
1893
+ dims: 512
1894
+ dims: 7
1895
+ data_type: 1
1896
+ name: "model.6.body.0.weight"
1897
+ }
1898
+ initializer {
1899
+ dims: 512
1900
+ data_type: 1
1901
+ name: "model.6.body.3.bias"
1902
+ }
1903
+ initializer {
1904
+ dims: 512
1905
+ dims: 512
1906
+ dims: 7
1907
+ data_type: 1
1908
+ name: "model.6.body.3.weight"
1909
+ }
1910
+ initializer {
1911
+ dims: 512
1912
+ data_type: 1
1913
+ name: "model.7.body.0.bias"
1914
+ }
1915
+ initializer {
1916
+ dims: 512
1917
+ dims: 512
1918
+ dims: 7
1919
+ data_type: 1
1920
+ name: "model.7.body.0.weight"
1921
+ }
1922
+ initializer {
1923
+ dims: 512
1924
+ data_type: 1
1925
+ name: "model.7.body.3.bias"
1926
+ }
1927
+ initializer {
1928
+ dims: 512
1929
+ dims: 512
1930
+ dims: 7
1931
+ data_type: 1
1932
+ name: "model.7.body.3.weight"
1933
+ }
1934
+ initializer {
1935
+ dims: 512
1936
+ data_type: 1
1937
+ name: "model.8.body.0.bias"
1938
+ }
1939
+ initializer {
1940
+ dims: 512
1941
+ dims: 512
1942
+ dims: 7
1943
+ data_type: 1
1944
+ name: "model.8.body.0.weight"
1945
+ }
1946
+ initializer {
1947
+ dims: 512
1948
+ data_type: 1
1949
+ name: "model.8.body.3.bias"
1950
+ }
1951
+ initializer {
1952
+ dims: 512
1953
+ dims: 512
1954
+ dims: 7
1955
+ data_type: 1
1956
+ name: "model.8.body.3.weight"
1957
+ }
1958
+ initializer {
1959
+ dims: 512
1960
+ data_type: 1
1961
+ name: "model.9.body.0.bias"
1962
+ }
1963
+ initializer {
1964
+ dims: 512
1965
+ dims: 512
1966
+ dims: 7
1967
+ data_type: 1
1968
+ name: "model.9.body.0.weight"
1969
+ }
1970
+ initializer {
1971
+ dims: 512
1972
+ data_type: 1
1973
+ name: "model.9.body.3.bias"
1974
+ }
1975
+ initializer {
1976
+ dims: 512
1977
+ dims: 512
1978
+ dims: 7
1979
+ data_type: 1
1980
+ name: "model.9.body.3.weight"
1981
+ }
1982
+ input {
1983
+ name: "x"
1984
+ type {
1985
+ tensor_type {
1986
+ elem_type: 1
1987
+ shape {
1988
+ dim {
1989
+ dim_param: "batch_size"
1990
+ }
1991
+ dim {
1992
+ dim_value: 2
1993
+ }
1994
+ dim {
1995
+ dim_value: 8192
1996
+ }
1997
+ }
1998
+ }
1999
+ }
2000
+ }
2001
+ output {
2002
+ name: "y"
2003
+ type {
2004
+ tensor_type {
2005
+ elem_type: 1
2006
+ shape {
2007
+ dim {
2008
+ dim_param: "batch_size"
2009
+ }
2010
+ dim {
2011
+ dim_value: 2
2012
+ }
2013
+ dim {
2014
+ dim_value: 8192
2015
+ }
2016
+ }
2017
+ }
2018
+ }
2019
+ }
2020
+ }
2021
+ opset_import {
2022
+ version: 11
2023
+ }
models/ailia-models/source.txt ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ https://github.com/axinc-ai/ailia-models/blob/master/audio_processing/deep-music-enhancer
2
+
3
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnet.onnx
4
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnet.onnx.prototxt
5
+
6
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetbn.onnx
7
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetbn.onnx.prototxt
8
+
9
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetda.onnx
10
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetda.onnx.prototxt
11
+
12
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetdo.onnx
13
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/resnetdo.onnx.prototxt
14
+
15
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unet.onnx
16
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unet.onnx.prototxt
17
+
18
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetbn.onnx
19
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetbn.onnx.prototxt
20
+
21
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetda.onnx
22
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetda.onnx.prototxt
23
+
24
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetdo.onnx
25
+ https://storage.googleapis.com/ailia-models/deep-music-enhancer/unetdo.onnx.prototxt
models/ailia-models/unet.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7d74cac5a843c3605b60a51880585b432ecc76259e188b50f0678d26fa55694
3
+ size 225706017
models/ailia-models/unet.onnx.prototxt ADDED
@@ -0,0 +1,2028 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "down_net.0.0.weight"
10
+ input: "down_net.0.0.bias"
11
+ output: "21"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 65
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 32
32
+ ints: 32
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 2
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "21"
43
+ output: "22"
44
+ name: "Relu_1"
45
+ op_type: "Relu"
46
+ }
47
+ node {
48
+ input: "22"
49
+ input: "down_net.1.0.weight"
50
+ input: "down_net.1.0.bias"
51
+ output: "23"
52
+ name: "Conv_2"
53
+ op_type: "Conv"
54
+ attribute {
55
+ name: "dilations"
56
+ ints: 1
57
+ type: INTS
58
+ }
59
+ attribute {
60
+ name: "group"
61
+ i: 1
62
+ type: INT
63
+ }
64
+ attribute {
65
+ name: "kernel_shape"
66
+ ints: 33
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "pads"
71
+ ints: 16
72
+ ints: 16
73
+ type: INTS
74
+ }
75
+ attribute {
76
+ name: "strides"
77
+ ints: 2
78
+ type: INTS
79
+ }
80
+ }
81
+ node {
82
+ input: "23"
83
+ output: "24"
84
+ name: "Relu_3"
85
+ op_type: "Relu"
86
+ }
87
+ node {
88
+ input: "24"
89
+ input: "down_net.2.0.weight"
90
+ input: "down_net.2.0.bias"
91
+ output: "25"
92
+ name: "Conv_4"
93
+ op_type: "Conv"
94
+ attribute {
95
+ name: "dilations"
96
+ ints: 1
97
+ type: INTS
98
+ }
99
+ attribute {
100
+ name: "group"
101
+ i: 1
102
+ type: INT
103
+ }
104
+ attribute {
105
+ name: "kernel_shape"
106
+ ints: 17
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "pads"
111
+ ints: 8
112
+ ints: 8
113
+ type: INTS
114
+ }
115
+ attribute {
116
+ name: "strides"
117
+ ints: 2
118
+ type: INTS
119
+ }
120
+ }
121
+ node {
122
+ input: "25"
123
+ output: "26"
124
+ name: "Relu_5"
125
+ op_type: "Relu"
126
+ }
127
+ node {
128
+ input: "26"
129
+ input: "down_net.3.0.weight"
130
+ input: "down_net.3.0.bias"
131
+ output: "27"
132
+ name: "Conv_6"
133
+ op_type: "Conv"
134
+ attribute {
135
+ name: "dilations"
136
+ ints: 1
137
+ type: INTS
138
+ }
139
+ attribute {
140
+ name: "group"
141
+ i: 1
142
+ type: INT
143
+ }
144
+ attribute {
145
+ name: "kernel_shape"
146
+ ints: 9
147
+ type: INTS
148
+ }
149
+ attribute {
150
+ name: "pads"
151
+ ints: 4
152
+ ints: 4
153
+ type: INTS
154
+ }
155
+ attribute {
156
+ name: "strides"
157
+ ints: 2
158
+ type: INTS
159
+ }
160
+ }
161
+ node {
162
+ input: "27"
163
+ output: "28"
164
+ name: "Relu_7"
165
+ op_type: "Relu"
166
+ }
167
+ node {
168
+ input: "28"
169
+ input: "bottleneck.0.weight"
170
+ input: "bottleneck.0.bias"
171
+ output: "29"
172
+ name: "Conv_8"
173
+ op_type: "Conv"
174
+ attribute {
175
+ name: "dilations"
176
+ ints: 1
177
+ type: INTS
178
+ }
179
+ attribute {
180
+ name: "group"
181
+ i: 1
182
+ type: INT
183
+ }
184
+ attribute {
185
+ name: "kernel_shape"
186
+ ints: 9
187
+ type: INTS
188
+ }
189
+ attribute {
190
+ name: "pads"
191
+ ints: 4
192
+ ints: 4
193
+ type: INTS
194
+ }
195
+ attribute {
196
+ name: "strides"
197
+ ints: 2
198
+ type: INTS
199
+ }
200
+ }
201
+ node {
202
+ input: "29"
203
+ output: "30"
204
+ name: "Relu_9"
205
+ op_type: "Relu"
206
+ }
207
+ node {
208
+ input: "30"
209
+ input: "bottleneck.2.weight"
210
+ input: "bottleneck.2.bias"
211
+ output: "31"
212
+ name: "Conv_10"
213
+ op_type: "Conv"
214
+ attribute {
215
+ name: "dilations"
216
+ ints: 1
217
+ type: INTS
218
+ }
219
+ attribute {
220
+ name: "group"
221
+ i: 1
222
+ type: INT
223
+ }
224
+ attribute {
225
+ name: "kernel_shape"
226
+ ints: 9
227
+ type: INTS
228
+ }
229
+ attribute {
230
+ name: "pads"
231
+ ints: 4
232
+ ints: 4
233
+ type: INTS
234
+ }
235
+ attribute {
236
+ name: "strides"
237
+ ints: 1
238
+ type: INTS
239
+ }
240
+ }
241
+ node {
242
+ input: "31"
243
+ output: "32"
244
+ name: "Relu_11"
245
+ op_type: "Relu"
246
+ }
247
+ node {
248
+ input: "32"
249
+ output: "33"
250
+ name: "Shape_12"
251
+ op_type: "Shape"
252
+ }
253
+ node {
254
+ output: "34"
255
+ name: "Constant_13"
256
+ op_type: "Constant"
257
+ attribute {
258
+ name: "value"
259
+ t {
260
+ data_type: 7
261
+ raw_data: "\000\000\000\000\000\000\000\000"
262
+ }
263
+ type: TENSOR
264
+ }
265
+ }
266
+ node {
267
+ input: "33"
268
+ input: "34"
269
+ output: "35"
270
+ name: "Gather_14"
271
+ op_type: "Gather"
272
+ attribute {
273
+ name: "axis"
274
+ i: 0
275
+ type: INT
276
+ }
277
+ }
278
+ node {
279
+ input: "32"
280
+ output: "36"
281
+ name: "Shape_15"
282
+ op_type: "Shape"
283
+ }
284
+ node {
285
+ output: "37"
286
+ name: "Constant_16"
287
+ op_type: "Constant"
288
+ attribute {
289
+ name: "value"
290
+ t {
291
+ data_type: 7
292
+ raw_data: "\001\000\000\000\000\000\000\000"
293
+ }
294
+ type: TENSOR
295
+ }
296
+ }
297
+ node {
298
+ input: "36"
299
+ input: "37"
300
+ output: "38"
301
+ name: "Gather_17"
302
+ op_type: "Gather"
303
+ attribute {
304
+ name: "axis"
305
+ i: 0
306
+ type: INT
307
+ }
308
+ }
309
+ node {
310
+ input: "32"
311
+ output: "39"
312
+ name: "Shape_18"
313
+ op_type: "Shape"
314
+ }
315
+ node {
316
+ output: "40"
317
+ name: "Constant_19"
318
+ op_type: "Constant"
319
+ attribute {
320
+ name: "value"
321
+ t {
322
+ data_type: 7
323
+ raw_data: "\002\000\000\000\000\000\000\000"
324
+ }
325
+ type: TENSOR
326
+ }
327
+ }
328
+ node {
329
+ input: "39"
330
+ input: "40"
331
+ output: "41"
332
+ name: "Gather_20"
333
+ op_type: "Gather"
334
+ attribute {
335
+ name: "axis"
336
+ i: 0
337
+ type: INT
338
+ }
339
+ }
340
+ node {
341
+ output: "42"
342
+ name: "Constant_21"
343
+ op_type: "Constant"
344
+ attribute {
345
+ name: "value"
346
+ t {
347
+ data_type: 7
348
+ raw_data: "\002\000\000\000\000\000\000\000"
349
+ }
350
+ type: TENSOR
351
+ }
352
+ }
353
+ node {
354
+ input: "38"
355
+ input: "42"
356
+ output: "43"
357
+ name: "Div_22"
358
+ op_type: "Div"
359
+ }
360
+ node {
361
+ input: "43"
362
+ output: "44"
363
+ name: "Cast_23"
364
+ op_type: "Cast"
365
+ attribute {
366
+ name: "to"
367
+ i: 7
368
+ type: INT
369
+ }
370
+ }
371
+ node {
372
+ input: "44"
373
+ output: "45"
374
+ name: "Cast_24"
375
+ op_type: "Cast"
376
+ attribute {
377
+ name: "to"
378
+ i: 7
379
+ type: INT
380
+ }
381
+ }
382
+ node {
383
+ output: "46"
384
+ name: "Constant_25"
385
+ op_type: "Constant"
386
+ attribute {
387
+ name: "value"
388
+ t {
389
+ data_type: 7
390
+ raw_data: "\002\000\000\000\000\000\000\000"
391
+ }
392
+ type: TENSOR
393
+ }
394
+ }
395
+ node {
396
+ input: "41"
397
+ input: "46"
398
+ output: "47"
399
+ name: "Mul_26"
400
+ op_type: "Mul"
401
+ }
402
+ node {
403
+ input: "35"
404
+ output: "49"
405
+ name: "Unsqueeze_27"
406
+ op_type: "Unsqueeze"
407
+ attribute {
408
+ name: "axes"
409
+ ints: 0
410
+ type: INTS
411
+ }
412
+ }
413
+ node {
414
+ input: "45"
415
+ output: "51"
416
+ name: "Unsqueeze_28"
417
+ op_type: "Unsqueeze"
418
+ attribute {
419
+ name: "axes"
420
+ ints: 0
421
+ type: INTS
422
+ }
423
+ }
424
+ node {
425
+ input: "41"
426
+ output: "52"
427
+ name: "Unsqueeze_29"
428
+ op_type: "Unsqueeze"
429
+ attribute {
430
+ name: "axes"
431
+ ints: 0
432
+ type: INTS
433
+ }
434
+ }
435
+ node {
436
+ input: "49"
437
+ input: "185"
438
+ input: "51"
439
+ input: "52"
440
+ output: "53"
441
+ name: "Concat_30"
442
+ op_type: "Concat"
443
+ attribute {
444
+ name: "axis"
445
+ i: 0
446
+ type: INT
447
+ }
448
+ }
449
+ node {
450
+ input: "32"
451
+ input: "53"
452
+ output: "54"
453
+ name: "Reshape_31"
454
+ op_type: "Reshape"
455
+ }
456
+ node {
457
+ input: "54"
458
+ output: "55"
459
+ name: "Transpose_32"
460
+ op_type: "Transpose"
461
+ attribute {
462
+ name: "perm"
463
+ ints: 0
464
+ ints: 2
465
+ ints: 3
466
+ ints: 1
467
+ type: INTS
468
+ }
469
+ }
470
+ node {
471
+ input: "35"
472
+ output: "56"
473
+ name: "Unsqueeze_33"
474
+ op_type: "Unsqueeze"
475
+ attribute {
476
+ name: "axes"
477
+ ints: 0
478
+ type: INTS
479
+ }
480
+ }
481
+ node {
482
+ input: "45"
483
+ output: "57"
484
+ name: "Unsqueeze_34"
485
+ op_type: "Unsqueeze"
486
+ attribute {
487
+ name: "axes"
488
+ ints: 0
489
+ type: INTS
490
+ }
491
+ }
492
+ node {
493
+ input: "47"
494
+ output: "58"
495
+ name: "Unsqueeze_35"
496
+ op_type: "Unsqueeze"
497
+ attribute {
498
+ name: "axes"
499
+ ints: 0
500
+ type: INTS
501
+ }
502
+ }
503
+ node {
504
+ input: "56"
505
+ input: "57"
506
+ input: "58"
507
+ output: "59"
508
+ name: "Concat_36"
509
+ op_type: "Concat"
510
+ attribute {
511
+ name: "axis"
512
+ i: 0
513
+ type: INT
514
+ }
515
+ }
516
+ node {
517
+ input: "55"
518
+ input: "59"
519
+ output: "60"
520
+ name: "Reshape_37"
521
+ op_type: "Reshape"
522
+ }
523
+ node {
524
+ input: "60"
525
+ input: "28"
526
+ output: "61"
527
+ name: "Concat_38"
528
+ op_type: "Concat"
529
+ attribute {
530
+ name: "axis"
531
+ i: 1
532
+ type: INT
533
+ }
534
+ }
535
+ node {
536
+ input: "61"
537
+ input: "up_net.0.0.weight"
538
+ input: "up_net.0.0.bias"
539
+ output: "62"
540
+ name: "Conv_39"
541
+ op_type: "Conv"
542
+ attribute {
543
+ name: "dilations"
544
+ ints: 1
545
+ type: INTS
546
+ }
547
+ attribute {
548
+ name: "group"
549
+ i: 1
550
+ type: INT
551
+ }
552
+ attribute {
553
+ name: "kernel_shape"
554
+ ints: 17
555
+ type: INTS
556
+ }
557
+ attribute {
558
+ name: "pads"
559
+ ints: 8
560
+ ints: 8
561
+ type: INTS
562
+ }
563
+ attribute {
564
+ name: "strides"
565
+ ints: 1
566
+ type: INTS
567
+ }
568
+ }
569
+ node {
570
+ input: "62"
571
+ output: "63"
572
+ name: "Relu_40"
573
+ op_type: "Relu"
574
+ }
575
+ node {
576
+ input: "63"
577
+ output: "64"
578
+ name: "Shape_41"
579
+ op_type: "Shape"
580
+ }
581
+ node {
582
+ output: "65"
583
+ name: "Constant_42"
584
+ op_type: "Constant"
585
+ attribute {
586
+ name: "value"
587
+ t {
588
+ data_type: 7
589
+ raw_data: "\000\000\000\000\000\000\000\000"
590
+ }
591
+ type: TENSOR
592
+ }
593
+ }
594
+ node {
595
+ input: "64"
596
+ input: "65"
597
+ output: "66"
598
+ name: "Gather_43"
599
+ op_type: "Gather"
600
+ attribute {
601
+ name: "axis"
602
+ i: 0
603
+ type: INT
604
+ }
605
+ }
606
+ node {
607
+ input: "63"
608
+ output: "67"
609
+ name: "Shape_44"
610
+ op_type: "Shape"
611
+ }
612
+ node {
613
+ output: "68"
614
+ name: "Constant_45"
615
+ op_type: "Constant"
616
+ attribute {
617
+ name: "value"
618
+ t {
619
+ data_type: 7
620
+ raw_data: "\001\000\000\000\000\000\000\000"
621
+ }
622
+ type: TENSOR
623
+ }
624
+ }
625
+ node {
626
+ input: "67"
627
+ input: "68"
628
+ output: "69"
629
+ name: "Gather_46"
630
+ op_type: "Gather"
631
+ attribute {
632
+ name: "axis"
633
+ i: 0
634
+ type: INT
635
+ }
636
+ }
637
+ node {
638
+ input: "63"
639
+ output: "70"
640
+ name: "Shape_47"
641
+ op_type: "Shape"
642
+ }
643
+ node {
644
+ output: "71"
645
+ name: "Constant_48"
646
+ op_type: "Constant"
647
+ attribute {
648
+ name: "value"
649
+ t {
650
+ data_type: 7
651
+ raw_data: "\002\000\000\000\000\000\000\000"
652
+ }
653
+ type: TENSOR
654
+ }
655
+ }
656
+ node {
657
+ input: "70"
658
+ input: "71"
659
+ output: "72"
660
+ name: "Gather_49"
661
+ op_type: "Gather"
662
+ attribute {
663
+ name: "axis"
664
+ i: 0
665
+ type: INT
666
+ }
667
+ }
668
+ node {
669
+ output: "73"
670
+ name: "Constant_50"
671
+ op_type: "Constant"
672
+ attribute {
673
+ name: "value"
674
+ t {
675
+ data_type: 7
676
+ raw_data: "\002\000\000\000\000\000\000\000"
677
+ }
678
+ type: TENSOR
679
+ }
680
+ }
681
+ node {
682
+ input: "69"
683
+ input: "73"
684
+ output: "74"
685
+ name: "Div_51"
686
+ op_type: "Div"
687
+ }
688
+ node {
689
+ input: "74"
690
+ output: "75"
691
+ name: "Cast_52"
692
+ op_type: "Cast"
693
+ attribute {
694
+ name: "to"
695
+ i: 7
696
+ type: INT
697
+ }
698
+ }
699
+ node {
700
+ input: "75"
701
+ output: "76"
702
+ name: "Cast_53"
703
+ op_type: "Cast"
704
+ attribute {
705
+ name: "to"
706
+ i: 7
707
+ type: INT
708
+ }
709
+ }
710
+ node {
711
+ output: "77"
712
+ name: "Constant_54"
713
+ op_type: "Constant"
714
+ attribute {
715
+ name: "value"
716
+ t {
717
+ data_type: 7
718
+ raw_data: "\002\000\000\000\000\000\000\000"
719
+ }
720
+ type: TENSOR
721
+ }
722
+ }
723
+ node {
724
+ input: "72"
725
+ input: "77"
726
+ output: "78"
727
+ name: "Mul_55"
728
+ op_type: "Mul"
729
+ }
730
+ node {
731
+ input: "66"
732
+ output: "80"
733
+ name: "Unsqueeze_56"
734
+ op_type: "Unsqueeze"
735
+ attribute {
736
+ name: "axes"
737
+ ints: 0
738
+ type: INTS
739
+ }
740
+ }
741
+ node {
742
+ input: "76"
743
+ output: "82"
744
+ name: "Unsqueeze_57"
745
+ op_type: "Unsqueeze"
746
+ attribute {
747
+ name: "axes"
748
+ ints: 0
749
+ type: INTS
750
+ }
751
+ }
752
+ node {
753
+ input: "72"
754
+ output: "83"
755
+ name: "Unsqueeze_58"
756
+ op_type: "Unsqueeze"
757
+ attribute {
758
+ name: "axes"
759
+ ints: 0
760
+ type: INTS
761
+ }
762
+ }
763
+ node {
764
+ input: "80"
765
+ input: "186"
766
+ input: "82"
767
+ input: "83"
768
+ output: "84"
769
+ name: "Concat_59"
770
+ op_type: "Concat"
771
+ attribute {
772
+ name: "axis"
773
+ i: 0
774
+ type: INT
775
+ }
776
+ }
777
+ node {
778
+ input: "63"
779
+ input: "84"
780
+ output: "85"
781
+ name: "Reshape_60"
782
+ op_type: "Reshape"
783
+ }
784
+ node {
785
+ input: "85"
786
+ output: "86"
787
+ name: "Transpose_61"
788
+ op_type: "Transpose"
789
+ attribute {
790
+ name: "perm"
791
+ ints: 0
792
+ ints: 2
793
+ ints: 3
794
+ ints: 1
795
+ type: INTS
796
+ }
797
+ }
798
+ node {
799
+ input: "66"
800
+ output: "87"
801
+ name: "Unsqueeze_62"
802
+ op_type: "Unsqueeze"
803
+ attribute {
804
+ name: "axes"
805
+ ints: 0
806
+ type: INTS
807
+ }
808
+ }
809
+ node {
810
+ input: "76"
811
+ output: "88"
812
+ name: "Unsqueeze_63"
813
+ op_type: "Unsqueeze"
814
+ attribute {
815
+ name: "axes"
816
+ ints: 0
817
+ type: INTS
818
+ }
819
+ }
820
+ node {
821
+ input: "78"
822
+ output: "89"
823
+ name: "Unsqueeze_64"
824
+ op_type: "Unsqueeze"
825
+ attribute {
826
+ name: "axes"
827
+ ints: 0
828
+ type: INTS
829
+ }
830
+ }
831
+ node {
832
+ input: "87"
833
+ input: "88"
834
+ input: "89"
835
+ output: "90"
836
+ name: "Concat_65"
837
+ op_type: "Concat"
838
+ attribute {
839
+ name: "axis"
840
+ i: 0
841
+ type: INT
842
+ }
843
+ }
844
+ node {
845
+ input: "86"
846
+ input: "90"
847
+ output: "91"
848
+ name: "Reshape_66"
849
+ op_type: "Reshape"
850
+ }
851
+ node {
852
+ input: "91"
853
+ input: "26"
854
+ output: "92"
855
+ name: "Concat_67"
856
+ op_type: "Concat"
857
+ attribute {
858
+ name: "axis"
859
+ i: 1
860
+ type: INT
861
+ }
862
+ }
863
+ node {
864
+ input: "92"
865
+ input: "up_net.1.0.weight"
866
+ input: "up_net.1.0.bias"
867
+ output: "93"
868
+ name: "Conv_68"
869
+ op_type: "Conv"
870
+ attribute {
871
+ name: "dilations"
872
+ ints: 1
873
+ type: INTS
874
+ }
875
+ attribute {
876
+ name: "group"
877
+ i: 1
878
+ type: INT
879
+ }
880
+ attribute {
881
+ name: "kernel_shape"
882
+ ints: 33
883
+ type: INTS
884
+ }
885
+ attribute {
886
+ name: "pads"
887
+ ints: 16
888
+ ints: 16
889
+ type: INTS
890
+ }
891
+ attribute {
892
+ name: "strides"
893
+ ints: 1
894
+ type: INTS
895
+ }
896
+ }
897
+ node {
898
+ input: "93"
899
+ output: "94"
900
+ name: "Relu_69"
901
+ op_type: "Relu"
902
+ }
903
+ node {
904
+ input: "94"
905
+ output: "95"
906
+ name: "Shape_70"
907
+ op_type: "Shape"
908
+ }
909
+ node {
910
+ output: "96"
911
+ name: "Constant_71"
912
+ op_type: "Constant"
913
+ attribute {
914
+ name: "value"
915
+ t {
916
+ data_type: 7
917
+ raw_data: "\000\000\000\000\000\000\000\000"
918
+ }
919
+ type: TENSOR
920
+ }
921
+ }
922
+ node {
923
+ input: "95"
924
+ input: "96"
925
+ output: "97"
926
+ name: "Gather_72"
927
+ op_type: "Gather"
928
+ attribute {
929
+ name: "axis"
930
+ i: 0
931
+ type: INT
932
+ }
933
+ }
934
+ node {
935
+ input: "94"
936
+ output: "98"
937
+ name: "Shape_73"
938
+ op_type: "Shape"
939
+ }
940
+ node {
941
+ output: "99"
942
+ name: "Constant_74"
943
+ op_type: "Constant"
944
+ attribute {
945
+ name: "value"
946
+ t {
947
+ data_type: 7
948
+ raw_data: "\001\000\000\000\000\000\000\000"
949
+ }
950
+ type: TENSOR
951
+ }
952
+ }
953
+ node {
954
+ input: "98"
955
+ input: "99"
956
+ output: "100"
957
+ name: "Gather_75"
958
+ op_type: "Gather"
959
+ attribute {
960
+ name: "axis"
961
+ i: 0
962
+ type: INT
963
+ }
964
+ }
965
+ node {
966
+ input: "94"
967
+ output: "101"
968
+ name: "Shape_76"
969
+ op_type: "Shape"
970
+ }
971
+ node {
972
+ output: "102"
973
+ name: "Constant_77"
974
+ op_type: "Constant"
975
+ attribute {
976
+ name: "value"
977
+ t {
978
+ data_type: 7
979
+ raw_data: "\002\000\000\000\000\000\000\000"
980
+ }
981
+ type: TENSOR
982
+ }
983
+ }
984
+ node {
985
+ input: "101"
986
+ input: "102"
987
+ output: "103"
988
+ name: "Gather_78"
989
+ op_type: "Gather"
990
+ attribute {
991
+ name: "axis"
992
+ i: 0
993
+ type: INT
994
+ }
995
+ }
996
+ node {
997
+ output: "104"
998
+ name: "Constant_79"
999
+ op_type: "Constant"
1000
+ attribute {
1001
+ name: "value"
1002
+ t {
1003
+ data_type: 7
1004
+ raw_data: "\002\000\000\000\000\000\000\000"
1005
+ }
1006
+ type: TENSOR
1007
+ }
1008
+ }
1009
+ node {
1010
+ input: "100"
1011
+ input: "104"
1012
+ output: "105"
1013
+ name: "Div_80"
1014
+ op_type: "Div"
1015
+ }
1016
+ node {
1017
+ input: "105"
1018
+ output: "106"
1019
+ name: "Cast_81"
1020
+ op_type: "Cast"
1021
+ attribute {
1022
+ name: "to"
1023
+ i: 7
1024
+ type: INT
1025
+ }
1026
+ }
1027
+ node {
1028
+ input: "106"
1029
+ output: "107"
1030
+ name: "Cast_82"
1031
+ op_type: "Cast"
1032
+ attribute {
1033
+ name: "to"
1034
+ i: 7
1035
+ type: INT
1036
+ }
1037
+ }
1038
+ node {
1039
+ output: "108"
1040
+ name: "Constant_83"
1041
+ op_type: "Constant"
1042
+ attribute {
1043
+ name: "value"
1044
+ t {
1045
+ data_type: 7
1046
+ raw_data: "\002\000\000\000\000\000\000\000"
1047
+ }
1048
+ type: TENSOR
1049
+ }
1050
+ }
1051
+ node {
1052
+ input: "103"
1053
+ input: "108"
1054
+ output: "109"
1055
+ name: "Mul_84"
1056
+ op_type: "Mul"
1057
+ }
1058
+ node {
1059
+ input: "97"
1060
+ output: "111"
1061
+ name: "Unsqueeze_85"
1062
+ op_type: "Unsqueeze"
1063
+ attribute {
1064
+ name: "axes"
1065
+ ints: 0
1066
+ type: INTS
1067
+ }
1068
+ }
1069
+ node {
1070
+ input: "107"
1071
+ output: "113"
1072
+ name: "Unsqueeze_86"
1073
+ op_type: "Unsqueeze"
1074
+ attribute {
1075
+ name: "axes"
1076
+ ints: 0
1077
+ type: INTS
1078
+ }
1079
+ }
1080
+ node {
1081
+ input: "103"
1082
+ output: "114"
1083
+ name: "Unsqueeze_87"
1084
+ op_type: "Unsqueeze"
1085
+ attribute {
1086
+ name: "axes"
1087
+ ints: 0
1088
+ type: INTS
1089
+ }
1090
+ }
1091
+ node {
1092
+ input: "111"
1093
+ input: "187"
1094
+ input: "113"
1095
+ input: "114"
1096
+ output: "115"
1097
+ name: "Concat_88"
1098
+ op_type: "Concat"
1099
+ attribute {
1100
+ name: "axis"
1101
+ i: 0
1102
+ type: INT
1103
+ }
1104
+ }
1105
+ node {
1106
+ input: "94"
1107
+ input: "115"
1108
+ output: "116"
1109
+ name: "Reshape_89"
1110
+ op_type: "Reshape"
1111
+ }
1112
+ node {
1113
+ input: "116"
1114
+ output: "117"
1115
+ name: "Transpose_90"
1116
+ op_type: "Transpose"
1117
+ attribute {
1118
+ name: "perm"
1119
+ ints: 0
1120
+ ints: 2
1121
+ ints: 3
1122
+ ints: 1
1123
+ type: INTS
1124
+ }
1125
+ }
1126
+ node {
1127
+ input: "97"
1128
+ output: "118"
1129
+ name: "Unsqueeze_91"
1130
+ op_type: "Unsqueeze"
1131
+ attribute {
1132
+ name: "axes"
1133
+ ints: 0
1134
+ type: INTS
1135
+ }
1136
+ }
1137
+ node {
1138
+ input: "107"
1139
+ output: "119"
1140
+ name: "Unsqueeze_92"
1141
+ op_type: "Unsqueeze"
1142
+ attribute {
1143
+ name: "axes"
1144
+ ints: 0
1145
+ type: INTS
1146
+ }
1147
+ }
1148
+ node {
1149
+ input: "109"
1150
+ output: "120"
1151
+ name: "Unsqueeze_93"
1152
+ op_type: "Unsqueeze"
1153
+ attribute {
1154
+ name: "axes"
1155
+ ints: 0
1156
+ type: INTS
1157
+ }
1158
+ }
1159
+ node {
1160
+ input: "118"
1161
+ input: "119"
1162
+ input: "120"
1163
+ output: "121"
1164
+ name: "Concat_94"
1165
+ op_type: "Concat"
1166
+ attribute {
1167
+ name: "axis"
1168
+ i: 0
1169
+ type: INT
1170
+ }
1171
+ }
1172
+ node {
1173
+ input: "117"
1174
+ input: "121"
1175
+ output: "122"
1176
+ name: "Reshape_95"
1177
+ op_type: "Reshape"
1178
+ }
1179
+ node {
1180
+ input: "122"
1181
+ input: "24"
1182
+ output: "123"
1183
+ name: "Concat_96"
1184
+ op_type: "Concat"
1185
+ attribute {
1186
+ name: "axis"
1187
+ i: 1
1188
+ type: INT
1189
+ }
1190
+ }
1191
+ node {
1192
+ input: "123"
1193
+ input: "up_net.2.0.weight"
1194
+ input: "up_net.2.0.bias"
1195
+ output: "124"
1196
+ name: "Conv_97"
1197
+ op_type: "Conv"
1198
+ attribute {
1199
+ name: "dilations"
1200
+ ints: 1
1201
+ type: INTS
1202
+ }
1203
+ attribute {
1204
+ name: "group"
1205
+ i: 1
1206
+ type: INT
1207
+ }
1208
+ attribute {
1209
+ name: "kernel_shape"
1210
+ ints: 65
1211
+ type: INTS
1212
+ }
1213
+ attribute {
1214
+ name: "pads"
1215
+ ints: 32
1216
+ ints: 32
1217
+ type: INTS
1218
+ }
1219
+ attribute {
1220
+ name: "strides"
1221
+ ints: 1
1222
+ type: INTS
1223
+ }
1224
+ }
1225
+ node {
1226
+ input: "124"
1227
+ output: "125"
1228
+ name: "Relu_98"
1229
+ op_type: "Relu"
1230
+ }
1231
+ node {
1232
+ input: "125"
1233
+ output: "126"
1234
+ name: "Shape_99"
1235
+ op_type: "Shape"
1236
+ }
1237
+ node {
1238
+ output: "127"
1239
+ name: "Constant_100"
1240
+ op_type: "Constant"
1241
+ attribute {
1242
+ name: "value"
1243
+ t {
1244
+ data_type: 7
1245
+ raw_data: "\000\000\000\000\000\000\000\000"
1246
+ }
1247
+ type: TENSOR
1248
+ }
1249
+ }
1250
+ node {
1251
+ input: "126"
1252
+ input: "127"
1253
+ output: "128"
1254
+ name: "Gather_101"
1255
+ op_type: "Gather"
1256
+ attribute {
1257
+ name: "axis"
1258
+ i: 0
1259
+ type: INT
1260
+ }
1261
+ }
1262
+ node {
1263
+ input: "125"
1264
+ output: "129"
1265
+ name: "Shape_102"
1266
+ op_type: "Shape"
1267
+ }
1268
+ node {
1269
+ output: "130"
1270
+ name: "Constant_103"
1271
+ op_type: "Constant"
1272
+ attribute {
1273
+ name: "value"
1274
+ t {
1275
+ data_type: 7
1276
+ raw_data: "\001\000\000\000\000\000\000\000"
1277
+ }
1278
+ type: TENSOR
1279
+ }
1280
+ }
1281
+ node {
1282
+ input: "129"
1283
+ input: "130"
1284
+ output: "131"
1285
+ name: "Gather_104"
1286
+ op_type: "Gather"
1287
+ attribute {
1288
+ name: "axis"
1289
+ i: 0
1290
+ type: INT
1291
+ }
1292
+ }
1293
+ node {
1294
+ input: "125"
1295
+ output: "132"
1296
+ name: "Shape_105"
1297
+ op_type: "Shape"
1298
+ }
1299
+ node {
1300
+ output: "133"
1301
+ name: "Constant_106"
1302
+ op_type: "Constant"
1303
+ attribute {
1304
+ name: "value"
1305
+ t {
1306
+ data_type: 7
1307
+ raw_data: "\002\000\000\000\000\000\000\000"
1308
+ }
1309
+ type: TENSOR
1310
+ }
1311
+ }
1312
+ node {
1313
+ input: "132"
1314
+ input: "133"
1315
+ output: "134"
1316
+ name: "Gather_107"
1317
+ op_type: "Gather"
1318
+ attribute {
1319
+ name: "axis"
1320
+ i: 0
1321
+ type: INT
1322
+ }
1323
+ }
1324
+ node {
1325
+ output: "135"
1326
+ name: "Constant_108"
1327
+ op_type: "Constant"
1328
+ attribute {
1329
+ name: "value"
1330
+ t {
1331
+ data_type: 7
1332
+ raw_data: "\002\000\000\000\000\000\000\000"
1333
+ }
1334
+ type: TENSOR
1335
+ }
1336
+ }
1337
+ node {
1338
+ input: "131"
1339
+ input: "135"
1340
+ output: "136"
1341
+ name: "Div_109"
1342
+ op_type: "Div"
1343
+ }
1344
+ node {
1345
+ input: "136"
1346
+ output: "137"
1347
+ name: "Cast_110"
1348
+ op_type: "Cast"
1349
+ attribute {
1350
+ name: "to"
1351
+ i: 7
1352
+ type: INT
1353
+ }
1354
+ }
1355
+ node {
1356
+ input: "137"
1357
+ output: "138"
1358
+ name: "Cast_111"
1359
+ op_type: "Cast"
1360
+ attribute {
1361
+ name: "to"
1362
+ i: 7
1363
+ type: INT
1364
+ }
1365
+ }
1366
+ node {
1367
+ output: "139"
1368
+ name: "Constant_112"
1369
+ op_type: "Constant"
1370
+ attribute {
1371
+ name: "value"
1372
+ t {
1373
+ data_type: 7
1374
+ raw_data: "\002\000\000\000\000\000\000\000"
1375
+ }
1376
+ type: TENSOR
1377
+ }
1378
+ }
1379
+ node {
1380
+ input: "134"
1381
+ input: "139"
1382
+ output: "140"
1383
+ name: "Mul_113"
1384
+ op_type: "Mul"
1385
+ }
1386
+ node {
1387
+ input: "128"
1388
+ output: "142"
1389
+ name: "Unsqueeze_114"
1390
+ op_type: "Unsqueeze"
1391
+ attribute {
1392
+ name: "axes"
1393
+ ints: 0
1394
+ type: INTS
1395
+ }
1396
+ }
1397
+ node {
1398
+ input: "138"
1399
+ output: "144"
1400
+ name: "Unsqueeze_115"
1401
+ op_type: "Unsqueeze"
1402
+ attribute {
1403
+ name: "axes"
1404
+ ints: 0
1405
+ type: INTS
1406
+ }
1407
+ }
1408
+ node {
1409
+ input: "134"
1410
+ output: "145"
1411
+ name: "Unsqueeze_116"
1412
+ op_type: "Unsqueeze"
1413
+ attribute {
1414
+ name: "axes"
1415
+ ints: 0
1416
+ type: INTS
1417
+ }
1418
+ }
1419
+ node {
1420
+ input: "142"
1421
+ input: "188"
1422
+ input: "144"
1423
+ input: "145"
1424
+ output: "146"
1425
+ name: "Concat_117"
1426
+ op_type: "Concat"
1427
+ attribute {
1428
+ name: "axis"
1429
+ i: 0
1430
+ type: INT
1431
+ }
1432
+ }
1433
+ node {
1434
+ input: "125"
1435
+ input: "146"
1436
+ output: "147"
1437
+ name: "Reshape_118"
1438
+ op_type: "Reshape"
1439
+ }
1440
+ node {
1441
+ input: "147"
1442
+ output: "148"
1443
+ name: "Transpose_119"
1444
+ op_type: "Transpose"
1445
+ attribute {
1446
+ name: "perm"
1447
+ ints: 0
1448
+ ints: 2
1449
+ ints: 3
1450
+ ints: 1
1451
+ type: INTS
1452
+ }
1453
+ }
1454
+ node {
1455
+ input: "128"
1456
+ output: "149"
1457
+ name: "Unsqueeze_120"
1458
+ op_type: "Unsqueeze"
1459
+ attribute {
1460
+ name: "axes"
1461
+ ints: 0
1462
+ type: INTS
1463
+ }
1464
+ }
1465
+ node {
1466
+ input: "138"
1467
+ output: "150"
1468
+ name: "Unsqueeze_121"
1469
+ op_type: "Unsqueeze"
1470
+ attribute {
1471
+ name: "axes"
1472
+ ints: 0
1473
+ type: INTS
1474
+ }
1475
+ }
1476
+ node {
1477
+ input: "140"
1478
+ output: "151"
1479
+ name: "Unsqueeze_122"
1480
+ op_type: "Unsqueeze"
1481
+ attribute {
1482
+ name: "axes"
1483
+ ints: 0
1484
+ type: INTS
1485
+ }
1486
+ }
1487
+ node {
1488
+ input: "149"
1489
+ input: "150"
1490
+ input: "151"
1491
+ output: "152"
1492
+ name: "Concat_123"
1493
+ op_type: "Concat"
1494
+ attribute {
1495
+ name: "axis"
1496
+ i: 0
1497
+ type: INT
1498
+ }
1499
+ }
1500
+ node {
1501
+ input: "148"
1502
+ input: "152"
1503
+ output: "153"
1504
+ name: "Reshape_124"
1505
+ op_type: "Reshape"
1506
+ }
1507
+ node {
1508
+ input: "153"
1509
+ input: "22"
1510
+ output: "154"
1511
+ name: "Concat_125"
1512
+ op_type: "Concat"
1513
+ attribute {
1514
+ name: "axis"
1515
+ i: 1
1516
+ type: INT
1517
+ }
1518
+ }
1519
+ node {
1520
+ input: "154"
1521
+ input: "up_net.3.0.weight"
1522
+ input: "up_net.3.0.bias"
1523
+ output: "155"
1524
+ name: "Conv_126"
1525
+ op_type: "Conv"
1526
+ attribute {
1527
+ name: "dilations"
1528
+ ints: 1
1529
+ type: INTS
1530
+ }
1531
+ attribute {
1532
+ name: "group"
1533
+ i: 1
1534
+ type: INT
1535
+ }
1536
+ attribute {
1537
+ name: "kernel_shape"
1538
+ ints: 9
1539
+ type: INTS
1540
+ }
1541
+ attribute {
1542
+ name: "pads"
1543
+ ints: 4
1544
+ ints: 4
1545
+ type: INTS
1546
+ }
1547
+ attribute {
1548
+ name: "strides"
1549
+ ints: 1
1550
+ type: INTS
1551
+ }
1552
+ }
1553
+ node {
1554
+ input: "155"
1555
+ output: "156"
1556
+ name: "Shape_127"
1557
+ op_type: "Shape"
1558
+ }
1559
+ node {
1560
+ output: "157"
1561
+ name: "Constant_128"
1562
+ op_type: "Constant"
1563
+ attribute {
1564
+ name: "value"
1565
+ t {
1566
+ data_type: 7
1567
+ raw_data: "\000\000\000\000\000\000\000\000"
1568
+ }
1569
+ type: TENSOR
1570
+ }
1571
+ }
1572
+ node {
1573
+ input: "156"
1574
+ input: "157"
1575
+ output: "158"
1576
+ name: "Gather_129"
1577
+ op_type: "Gather"
1578
+ attribute {
1579
+ name: "axis"
1580
+ i: 0
1581
+ type: INT
1582
+ }
1583
+ }
1584
+ node {
1585
+ input: "155"
1586
+ output: "159"
1587
+ name: "Shape_130"
1588
+ op_type: "Shape"
1589
+ }
1590
+ node {
1591
+ output: "160"
1592
+ name: "Constant_131"
1593
+ op_type: "Constant"
1594
+ attribute {
1595
+ name: "value"
1596
+ t {
1597
+ data_type: 7
1598
+ raw_data: "\001\000\000\000\000\000\000\000"
1599
+ }
1600
+ type: TENSOR
1601
+ }
1602
+ }
1603
+ node {
1604
+ input: "159"
1605
+ input: "160"
1606
+ output: "161"
1607
+ name: "Gather_132"
1608
+ op_type: "Gather"
1609
+ attribute {
1610
+ name: "axis"
1611
+ i: 0
1612
+ type: INT
1613
+ }
1614
+ }
1615
+ node {
1616
+ input: "155"
1617
+ output: "162"
1618
+ name: "Shape_133"
1619
+ op_type: "Shape"
1620
+ }
1621
+ node {
1622
+ output: "163"
1623
+ name: "Constant_134"
1624
+ op_type: "Constant"
1625
+ attribute {
1626
+ name: "value"
1627
+ t {
1628
+ data_type: 7
1629
+ raw_data: "\002\000\000\000\000\000\000\000"
1630
+ }
1631
+ type: TENSOR
1632
+ }
1633
+ }
1634
+ node {
1635
+ input: "162"
1636
+ input: "163"
1637
+ output: "164"
1638
+ name: "Gather_135"
1639
+ op_type: "Gather"
1640
+ attribute {
1641
+ name: "axis"
1642
+ i: 0
1643
+ type: INT
1644
+ }
1645
+ }
1646
+ node {
1647
+ output: "165"
1648
+ name: "Constant_136"
1649
+ op_type: "Constant"
1650
+ attribute {
1651
+ name: "value"
1652
+ t {
1653
+ data_type: 7
1654
+ raw_data: "\002\000\000\000\000\000\000\000"
1655
+ }
1656
+ type: TENSOR
1657
+ }
1658
+ }
1659
+ node {
1660
+ input: "161"
1661
+ input: "165"
1662
+ output: "166"
1663
+ name: "Div_137"
1664
+ op_type: "Div"
1665
+ }
1666
+ node {
1667
+ input: "166"
1668
+ output: "167"
1669
+ name: "Cast_138"
1670
+ op_type: "Cast"
1671
+ attribute {
1672
+ name: "to"
1673
+ i: 7
1674
+ type: INT
1675
+ }
1676
+ }
1677
+ node {
1678
+ input: "167"
1679
+ output: "168"
1680
+ name: "Cast_139"
1681
+ op_type: "Cast"
1682
+ attribute {
1683
+ name: "to"
1684
+ i: 7
1685
+ type: INT
1686
+ }
1687
+ }
1688
+ node {
1689
+ output: "169"
1690
+ name: "Constant_140"
1691
+ op_type: "Constant"
1692
+ attribute {
1693
+ name: "value"
1694
+ t {
1695
+ data_type: 7
1696
+ raw_data: "\002\000\000\000\000\000\000\000"
1697
+ }
1698
+ type: TENSOR
1699
+ }
1700
+ }
1701
+ node {
1702
+ input: "164"
1703
+ input: "169"
1704
+ output: "170"
1705
+ name: "Mul_141"
1706
+ op_type: "Mul"
1707
+ }
1708
+ node {
1709
+ input: "158"
1710
+ output: "172"
1711
+ name: "Unsqueeze_142"
1712
+ op_type: "Unsqueeze"
1713
+ attribute {
1714
+ name: "axes"
1715
+ ints: 0
1716
+ type: INTS
1717
+ }
1718
+ }
1719
+ node {
1720
+ input: "168"
1721
+ output: "174"
1722
+ name: "Unsqueeze_143"
1723
+ op_type: "Unsqueeze"
1724
+ attribute {
1725
+ name: "axes"
1726
+ ints: 0
1727
+ type: INTS
1728
+ }
1729
+ }
1730
+ node {
1731
+ input: "164"
1732
+ output: "175"
1733
+ name: "Unsqueeze_144"
1734
+ op_type: "Unsqueeze"
1735
+ attribute {
1736
+ name: "axes"
1737
+ ints: 0
1738
+ type: INTS
1739
+ }
1740
+ }
1741
+ node {
1742
+ input: "172"
1743
+ input: "189"
1744
+ input: "174"
1745
+ input: "175"
1746
+ output: "176"
1747
+ name: "Concat_145"
1748
+ op_type: "Concat"
1749
+ attribute {
1750
+ name: "axis"
1751
+ i: 0
1752
+ type: INT
1753
+ }
1754
+ }
1755
+ node {
1756
+ input: "155"
1757
+ input: "176"
1758
+ output: "177"
1759
+ name: "Reshape_146"
1760
+ op_type: "Reshape"
1761
+ }
1762
+ node {
1763
+ input: "177"
1764
+ output: "178"
1765
+ name: "Transpose_147"
1766
+ op_type: "Transpose"
1767
+ attribute {
1768
+ name: "perm"
1769
+ ints: 0
1770
+ ints: 2
1771
+ ints: 3
1772
+ ints: 1
1773
+ type: INTS
1774
+ }
1775
+ }
1776
+ node {
1777
+ input: "158"
1778
+ output: "179"
1779
+ name: "Unsqueeze_148"
1780
+ op_type: "Unsqueeze"
1781
+ attribute {
1782
+ name: "axes"
1783
+ ints: 0
1784
+ type: INTS
1785
+ }
1786
+ }
1787
+ node {
1788
+ input: "168"
1789
+ output: "180"
1790
+ name: "Unsqueeze_149"
1791
+ op_type: "Unsqueeze"
1792
+ attribute {
1793
+ name: "axes"
1794
+ ints: 0
1795
+ type: INTS
1796
+ }
1797
+ }
1798
+ node {
1799
+ input: "170"
1800
+ output: "181"
1801
+ name: "Unsqueeze_150"
1802
+ op_type: "Unsqueeze"
1803
+ attribute {
1804
+ name: "axes"
1805
+ ints: 0
1806
+ type: INTS
1807
+ }
1808
+ }
1809
+ node {
1810
+ input: "179"
1811
+ input: "180"
1812
+ input: "181"
1813
+ output: "182"
1814
+ name: "Concat_151"
1815
+ op_type: "Concat"
1816
+ attribute {
1817
+ name: "axis"
1818
+ i: 0
1819
+ type: INT
1820
+ }
1821
+ }
1822
+ node {
1823
+ input: "178"
1824
+ input: "182"
1825
+ output: "183"
1826
+ name: "Reshape_152"
1827
+ op_type: "Reshape"
1828
+ }
1829
+ node {
1830
+ input: "183"
1831
+ input: "x"
1832
+ output: "y"
1833
+ name: "Add_153"
1834
+ op_type: "Add"
1835
+ }
1836
+ initializer {
1837
+ dims: 1
1838
+ data_type: 7
1839
+ name: "185"
1840
+ raw_data: "\002\000\000\000\000\000\000\000"
1841
+ }
1842
+ initializer {
1843
+ dims: 1
1844
+ data_type: 7
1845
+ name: "186"
1846
+ raw_data: "\002\000\000\000\000\000\000\000"
1847
+ }
1848
+ initializer {
1849
+ dims: 1
1850
+ data_type: 7
1851
+ name: "187"
1852
+ raw_data: "\002\000\000\000\000\000\000\000"
1853
+ }
1854
+ initializer {
1855
+ dims: 1
1856
+ data_type: 7
1857
+ name: "188"
1858
+ raw_data: "\002\000\000\000\000\000\000\000"
1859
+ }
1860
+ initializer {
1861
+ dims: 1
1862
+ data_type: 7
1863
+ name: "189"
1864
+ raw_data: "\002\000\000\000\000\000\000\000"
1865
+ }
1866
+ initializer {
1867
+ dims: 512
1868
+ data_type: 1
1869
+ name: "bottleneck.0.bias"
1870
+ }
1871
+ initializer {
1872
+ dims: 512
1873
+ dims: 512
1874
+ dims: 9
1875
+ data_type: 1
1876
+ name: "bottleneck.0.weight"
1877
+ }
1878
+ initializer {
1879
+ dims: 1024
1880
+ data_type: 1
1881
+ name: "bottleneck.2.bias"
1882
+ }
1883
+ initializer {
1884
+ dims: 1024
1885
+ dims: 512
1886
+ dims: 9
1887
+ data_type: 1
1888
+ name: "bottleneck.2.weight"
1889
+ }
1890
+ initializer {
1891
+ dims: 128
1892
+ data_type: 1
1893
+ name: "down_net.0.0.bias"
1894
+ }
1895
+ initializer {
1896
+ dims: 128
1897
+ dims: 2
1898
+ dims: 65
1899
+ data_type: 1
1900
+ name: "down_net.0.0.weight"
1901
+ }
1902
+ initializer {
1903
+ dims: 256
1904
+ data_type: 1
1905
+ name: "down_net.1.0.bias"
1906
+ }
1907
+ initializer {
1908
+ dims: 256
1909
+ dims: 128
1910
+ dims: 33
1911
+ data_type: 1
1912
+ name: "down_net.1.0.weight"
1913
+ }
1914
+ initializer {
1915
+ dims: 512
1916
+ data_type: 1
1917
+ name: "down_net.2.0.bias"
1918
+ }
1919
+ initializer {
1920
+ dims: 512
1921
+ dims: 256
1922
+ dims: 17
1923
+ data_type: 1
1924
+ name: "down_net.2.0.weight"
1925
+ }
1926
+ initializer {
1927
+ dims: 512
1928
+ data_type: 1
1929
+ name: "down_net.3.0.bias"
1930
+ }
1931
+ initializer {
1932
+ dims: 512
1933
+ dims: 512
1934
+ dims: 9
1935
+ data_type: 1
1936
+ name: "down_net.3.0.weight"
1937
+ }
1938
+ initializer {
1939
+ dims: 1024
1940
+ data_type: 1
1941
+ name: "up_net.0.0.bias"
1942
+ }
1943
+ initializer {
1944
+ dims: 1024
1945
+ dims: 1024
1946
+ dims: 17
1947
+ data_type: 1
1948
+ name: "up_net.0.0.weight"
1949
+ }
1950
+ initializer {
1951
+ dims: 512
1952
+ data_type: 1
1953
+ name: "up_net.1.0.bias"
1954
+ }
1955
+ initializer {
1956
+ dims: 512
1957
+ dims: 1024
1958
+ dims: 33
1959
+ data_type: 1
1960
+ name: "up_net.1.0.weight"
1961
+ }
1962
+ initializer {
1963
+ dims: 256
1964
+ data_type: 1
1965
+ name: "up_net.2.0.bias"
1966
+ }
1967
+ initializer {
1968
+ dims: 256
1969
+ dims: 512
1970
+ dims: 65
1971
+ data_type: 1
1972
+ name: "up_net.2.0.weight"
1973
+ }
1974
+ initializer {
1975
+ dims: 4
1976
+ data_type: 1
1977
+ name: "up_net.3.0.bias"
1978
+ raw_data: "\024pn\274JfK\274:s\223<F\316\017:"
1979
+ }
1980
+ initializer {
1981
+ dims: 4
1982
+ dims: 256
1983
+ dims: 9
1984
+ data_type: 1
1985
+ name: "up_net.3.0.weight"
1986
+ }
1987
+ input {
1988
+ name: "x"
1989
+ type {
1990
+ tensor_type {
1991
+ elem_type: 1
1992
+ shape {
1993
+ dim {
1994
+ dim_param: "batch_size"
1995
+ }
1996
+ dim {
1997
+ dim_value: 2
1998
+ }
1999
+ dim {
2000
+ dim_value: 8192
2001
+ }
2002
+ }
2003
+ }
2004
+ }
2005
+ }
2006
+ output {
2007
+ name: "y"
2008
+ type {
2009
+ tensor_type {
2010
+ elem_type: 1
2011
+ shape {
2012
+ dim {
2013
+ dim_param: "batch_size"
2014
+ }
2015
+ dim {
2016
+ dim_value: 2
2017
+ }
2018
+ dim {
2019
+ dim_value: 8192
2020
+ }
2021
+ }
2022
+ }
2023
+ }
2024
+ }
2025
+ }
2026
+ opset_import {
2027
+ version: 11
2028
+ }
models/ailia-models/unetbn.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a46f281c534a934d463b29687b92bbfb158cd3bf54d18c4bd33fcf45275a35e0
3
+ size 225705830
models/ailia-models/unetbn.onnx.prototxt ADDED
@@ -0,0 +1,2028 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "210"
10
+ input: "211"
11
+ output: "209"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 65
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 32
32
+ ints: 32
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 2
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "209"
43
+ output: "43"
44
+ name: "Relu_1"
45
+ op_type: "Relu"
46
+ }
47
+ node {
48
+ input: "43"
49
+ input: "213"
50
+ input: "214"
51
+ output: "212"
52
+ name: "Conv_2"
53
+ op_type: "Conv"
54
+ attribute {
55
+ name: "dilations"
56
+ ints: 1
57
+ type: INTS
58
+ }
59
+ attribute {
60
+ name: "group"
61
+ i: 1
62
+ type: INT
63
+ }
64
+ attribute {
65
+ name: "kernel_shape"
66
+ ints: 33
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "pads"
71
+ ints: 16
72
+ ints: 16
73
+ type: INTS
74
+ }
75
+ attribute {
76
+ name: "strides"
77
+ ints: 2
78
+ type: INTS
79
+ }
80
+ }
81
+ node {
82
+ input: "212"
83
+ output: "46"
84
+ name: "Relu_3"
85
+ op_type: "Relu"
86
+ }
87
+ node {
88
+ input: "46"
89
+ input: "216"
90
+ input: "217"
91
+ output: "215"
92
+ name: "Conv_4"
93
+ op_type: "Conv"
94
+ attribute {
95
+ name: "dilations"
96
+ ints: 1
97
+ type: INTS
98
+ }
99
+ attribute {
100
+ name: "group"
101
+ i: 1
102
+ type: INT
103
+ }
104
+ attribute {
105
+ name: "kernel_shape"
106
+ ints: 17
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "pads"
111
+ ints: 8
112
+ ints: 8
113
+ type: INTS
114
+ }
115
+ attribute {
116
+ name: "strides"
117
+ ints: 2
118
+ type: INTS
119
+ }
120
+ }
121
+ node {
122
+ input: "215"
123
+ output: "49"
124
+ name: "Relu_5"
125
+ op_type: "Relu"
126
+ }
127
+ node {
128
+ input: "49"
129
+ input: "219"
130
+ input: "220"
131
+ output: "218"
132
+ name: "Conv_6"
133
+ op_type: "Conv"
134
+ attribute {
135
+ name: "dilations"
136
+ ints: 1
137
+ type: INTS
138
+ }
139
+ attribute {
140
+ name: "group"
141
+ i: 1
142
+ type: INT
143
+ }
144
+ attribute {
145
+ name: "kernel_shape"
146
+ ints: 9
147
+ type: INTS
148
+ }
149
+ attribute {
150
+ name: "pads"
151
+ ints: 4
152
+ ints: 4
153
+ type: INTS
154
+ }
155
+ attribute {
156
+ name: "strides"
157
+ ints: 2
158
+ type: INTS
159
+ }
160
+ }
161
+ node {
162
+ input: "218"
163
+ output: "52"
164
+ name: "Relu_7"
165
+ op_type: "Relu"
166
+ }
167
+ node {
168
+ input: "52"
169
+ input: "bottleneck.0.weight"
170
+ input: "bottleneck.0.bias"
171
+ output: "53"
172
+ name: "Conv_8"
173
+ op_type: "Conv"
174
+ attribute {
175
+ name: "dilations"
176
+ ints: 1
177
+ type: INTS
178
+ }
179
+ attribute {
180
+ name: "group"
181
+ i: 1
182
+ type: INT
183
+ }
184
+ attribute {
185
+ name: "kernel_shape"
186
+ ints: 9
187
+ type: INTS
188
+ }
189
+ attribute {
190
+ name: "pads"
191
+ ints: 4
192
+ ints: 4
193
+ type: INTS
194
+ }
195
+ attribute {
196
+ name: "strides"
197
+ ints: 2
198
+ type: INTS
199
+ }
200
+ }
201
+ node {
202
+ input: "53"
203
+ output: "54"
204
+ name: "Relu_9"
205
+ op_type: "Relu"
206
+ }
207
+ node {
208
+ input: "54"
209
+ input: "bottleneck.2.weight"
210
+ input: "bottleneck.2.bias"
211
+ output: "55"
212
+ name: "Conv_10"
213
+ op_type: "Conv"
214
+ attribute {
215
+ name: "dilations"
216
+ ints: 1
217
+ type: INTS
218
+ }
219
+ attribute {
220
+ name: "group"
221
+ i: 1
222
+ type: INT
223
+ }
224
+ attribute {
225
+ name: "kernel_shape"
226
+ ints: 9
227
+ type: INTS
228
+ }
229
+ attribute {
230
+ name: "pads"
231
+ ints: 4
232
+ ints: 4
233
+ type: INTS
234
+ }
235
+ attribute {
236
+ name: "strides"
237
+ ints: 1
238
+ type: INTS
239
+ }
240
+ }
241
+ node {
242
+ input: "55"
243
+ output: "56"
244
+ name: "Relu_11"
245
+ op_type: "Relu"
246
+ }
247
+ node {
248
+ input: "56"
249
+ output: "57"
250
+ name: "Shape_12"
251
+ op_type: "Shape"
252
+ }
253
+ node {
254
+ output: "58"
255
+ name: "Constant_13"
256
+ op_type: "Constant"
257
+ attribute {
258
+ name: "value"
259
+ t {
260
+ data_type: 7
261
+ raw_data: "\000\000\000\000\000\000\000\000"
262
+ }
263
+ type: TENSOR
264
+ }
265
+ }
266
+ node {
267
+ input: "57"
268
+ input: "58"
269
+ output: "59"
270
+ name: "Gather_14"
271
+ op_type: "Gather"
272
+ attribute {
273
+ name: "axis"
274
+ i: 0
275
+ type: INT
276
+ }
277
+ }
278
+ node {
279
+ input: "56"
280
+ output: "60"
281
+ name: "Shape_15"
282
+ op_type: "Shape"
283
+ }
284
+ node {
285
+ output: "61"
286
+ name: "Constant_16"
287
+ op_type: "Constant"
288
+ attribute {
289
+ name: "value"
290
+ t {
291
+ data_type: 7
292
+ raw_data: "\001\000\000\000\000\000\000\000"
293
+ }
294
+ type: TENSOR
295
+ }
296
+ }
297
+ node {
298
+ input: "60"
299
+ input: "61"
300
+ output: "62"
301
+ name: "Gather_17"
302
+ op_type: "Gather"
303
+ attribute {
304
+ name: "axis"
305
+ i: 0
306
+ type: INT
307
+ }
308
+ }
309
+ node {
310
+ input: "56"
311
+ output: "63"
312
+ name: "Shape_18"
313
+ op_type: "Shape"
314
+ }
315
+ node {
316
+ output: "64"
317
+ name: "Constant_19"
318
+ op_type: "Constant"
319
+ attribute {
320
+ name: "value"
321
+ t {
322
+ data_type: 7
323
+ raw_data: "\002\000\000\000\000\000\000\000"
324
+ }
325
+ type: TENSOR
326
+ }
327
+ }
328
+ node {
329
+ input: "63"
330
+ input: "64"
331
+ output: "65"
332
+ name: "Gather_20"
333
+ op_type: "Gather"
334
+ attribute {
335
+ name: "axis"
336
+ i: 0
337
+ type: INT
338
+ }
339
+ }
340
+ node {
341
+ output: "66"
342
+ name: "Constant_21"
343
+ op_type: "Constant"
344
+ attribute {
345
+ name: "value"
346
+ t {
347
+ data_type: 7
348
+ raw_data: "\002\000\000\000\000\000\000\000"
349
+ }
350
+ type: TENSOR
351
+ }
352
+ }
353
+ node {
354
+ input: "62"
355
+ input: "66"
356
+ output: "67"
357
+ name: "Div_22"
358
+ op_type: "Div"
359
+ }
360
+ node {
361
+ input: "67"
362
+ output: "68"
363
+ name: "Cast_23"
364
+ op_type: "Cast"
365
+ attribute {
366
+ name: "to"
367
+ i: 7
368
+ type: INT
369
+ }
370
+ }
371
+ node {
372
+ input: "68"
373
+ output: "69"
374
+ name: "Cast_24"
375
+ op_type: "Cast"
376
+ attribute {
377
+ name: "to"
378
+ i: 7
379
+ type: INT
380
+ }
381
+ }
382
+ node {
383
+ output: "70"
384
+ name: "Constant_25"
385
+ op_type: "Constant"
386
+ attribute {
387
+ name: "value"
388
+ t {
389
+ data_type: 7
390
+ raw_data: "\002\000\000\000\000\000\000\000"
391
+ }
392
+ type: TENSOR
393
+ }
394
+ }
395
+ node {
396
+ input: "65"
397
+ input: "70"
398
+ output: "71"
399
+ name: "Mul_26"
400
+ op_type: "Mul"
401
+ }
402
+ node {
403
+ input: "59"
404
+ output: "73"
405
+ name: "Unsqueeze_27"
406
+ op_type: "Unsqueeze"
407
+ attribute {
408
+ name: "axes"
409
+ ints: 0
410
+ type: INTS
411
+ }
412
+ }
413
+ node {
414
+ input: "69"
415
+ output: "75"
416
+ name: "Unsqueeze_28"
417
+ op_type: "Unsqueeze"
418
+ attribute {
419
+ name: "axes"
420
+ ints: 0
421
+ type: INTS
422
+ }
423
+ }
424
+ node {
425
+ input: "65"
426
+ output: "76"
427
+ name: "Unsqueeze_29"
428
+ op_type: "Unsqueeze"
429
+ attribute {
430
+ name: "axes"
431
+ ints: 0
432
+ type: INTS
433
+ }
434
+ }
435
+ node {
436
+ input: "73"
437
+ input: "221"
438
+ input: "75"
439
+ input: "76"
440
+ output: "77"
441
+ name: "Concat_30"
442
+ op_type: "Concat"
443
+ attribute {
444
+ name: "axis"
445
+ i: 0
446
+ type: INT
447
+ }
448
+ }
449
+ node {
450
+ input: "56"
451
+ input: "77"
452
+ output: "78"
453
+ name: "Reshape_31"
454
+ op_type: "Reshape"
455
+ }
456
+ node {
457
+ input: "78"
458
+ output: "79"
459
+ name: "Transpose_32"
460
+ op_type: "Transpose"
461
+ attribute {
462
+ name: "perm"
463
+ ints: 0
464
+ ints: 2
465
+ ints: 3
466
+ ints: 1
467
+ type: INTS
468
+ }
469
+ }
470
+ node {
471
+ input: "59"
472
+ output: "80"
473
+ name: "Unsqueeze_33"
474
+ op_type: "Unsqueeze"
475
+ attribute {
476
+ name: "axes"
477
+ ints: 0
478
+ type: INTS
479
+ }
480
+ }
481
+ node {
482
+ input: "69"
483
+ output: "81"
484
+ name: "Unsqueeze_34"
485
+ op_type: "Unsqueeze"
486
+ attribute {
487
+ name: "axes"
488
+ ints: 0
489
+ type: INTS
490
+ }
491
+ }
492
+ node {
493
+ input: "71"
494
+ output: "82"
495
+ name: "Unsqueeze_35"
496
+ op_type: "Unsqueeze"
497
+ attribute {
498
+ name: "axes"
499
+ ints: 0
500
+ type: INTS
501
+ }
502
+ }
503
+ node {
504
+ input: "80"
505
+ input: "81"
506
+ input: "82"
507
+ output: "83"
508
+ name: "Concat_36"
509
+ op_type: "Concat"
510
+ attribute {
511
+ name: "axis"
512
+ i: 0
513
+ type: INT
514
+ }
515
+ }
516
+ node {
517
+ input: "79"
518
+ input: "83"
519
+ output: "84"
520
+ name: "Reshape_37"
521
+ op_type: "Reshape"
522
+ }
523
+ node {
524
+ input: "84"
525
+ input: "52"
526
+ output: "85"
527
+ name: "Concat_38"
528
+ op_type: "Concat"
529
+ attribute {
530
+ name: "axis"
531
+ i: 1
532
+ type: INT
533
+ }
534
+ }
535
+ node {
536
+ input: "85"
537
+ input: "up_net.0.0.weight"
538
+ input: "up_net.0.0.bias"
539
+ output: "86"
540
+ name: "Conv_39"
541
+ op_type: "Conv"
542
+ attribute {
543
+ name: "dilations"
544
+ ints: 1
545
+ type: INTS
546
+ }
547
+ attribute {
548
+ name: "group"
549
+ i: 1
550
+ type: INT
551
+ }
552
+ attribute {
553
+ name: "kernel_shape"
554
+ ints: 17
555
+ type: INTS
556
+ }
557
+ attribute {
558
+ name: "pads"
559
+ ints: 8
560
+ ints: 8
561
+ type: INTS
562
+ }
563
+ attribute {
564
+ name: "strides"
565
+ ints: 1
566
+ type: INTS
567
+ }
568
+ }
569
+ node {
570
+ input: "86"
571
+ output: "87"
572
+ name: "Relu_40"
573
+ op_type: "Relu"
574
+ }
575
+ node {
576
+ input: "87"
577
+ output: "88"
578
+ name: "Shape_41"
579
+ op_type: "Shape"
580
+ }
581
+ node {
582
+ output: "89"
583
+ name: "Constant_42"
584
+ op_type: "Constant"
585
+ attribute {
586
+ name: "value"
587
+ t {
588
+ data_type: 7
589
+ raw_data: "\000\000\000\000\000\000\000\000"
590
+ }
591
+ type: TENSOR
592
+ }
593
+ }
594
+ node {
595
+ input: "88"
596
+ input: "89"
597
+ output: "90"
598
+ name: "Gather_43"
599
+ op_type: "Gather"
600
+ attribute {
601
+ name: "axis"
602
+ i: 0
603
+ type: INT
604
+ }
605
+ }
606
+ node {
607
+ input: "87"
608
+ output: "91"
609
+ name: "Shape_44"
610
+ op_type: "Shape"
611
+ }
612
+ node {
613
+ output: "92"
614
+ name: "Constant_45"
615
+ op_type: "Constant"
616
+ attribute {
617
+ name: "value"
618
+ t {
619
+ data_type: 7
620
+ raw_data: "\001\000\000\000\000\000\000\000"
621
+ }
622
+ type: TENSOR
623
+ }
624
+ }
625
+ node {
626
+ input: "91"
627
+ input: "92"
628
+ output: "93"
629
+ name: "Gather_46"
630
+ op_type: "Gather"
631
+ attribute {
632
+ name: "axis"
633
+ i: 0
634
+ type: INT
635
+ }
636
+ }
637
+ node {
638
+ input: "87"
639
+ output: "94"
640
+ name: "Shape_47"
641
+ op_type: "Shape"
642
+ }
643
+ node {
644
+ output: "95"
645
+ name: "Constant_48"
646
+ op_type: "Constant"
647
+ attribute {
648
+ name: "value"
649
+ t {
650
+ data_type: 7
651
+ raw_data: "\002\000\000\000\000\000\000\000"
652
+ }
653
+ type: TENSOR
654
+ }
655
+ }
656
+ node {
657
+ input: "94"
658
+ input: "95"
659
+ output: "96"
660
+ name: "Gather_49"
661
+ op_type: "Gather"
662
+ attribute {
663
+ name: "axis"
664
+ i: 0
665
+ type: INT
666
+ }
667
+ }
668
+ node {
669
+ output: "97"
670
+ name: "Constant_50"
671
+ op_type: "Constant"
672
+ attribute {
673
+ name: "value"
674
+ t {
675
+ data_type: 7
676
+ raw_data: "\002\000\000\000\000\000\000\000"
677
+ }
678
+ type: TENSOR
679
+ }
680
+ }
681
+ node {
682
+ input: "93"
683
+ input: "97"
684
+ output: "98"
685
+ name: "Div_51"
686
+ op_type: "Div"
687
+ }
688
+ node {
689
+ input: "98"
690
+ output: "99"
691
+ name: "Cast_52"
692
+ op_type: "Cast"
693
+ attribute {
694
+ name: "to"
695
+ i: 7
696
+ type: INT
697
+ }
698
+ }
699
+ node {
700
+ input: "99"
701
+ output: "100"
702
+ name: "Cast_53"
703
+ op_type: "Cast"
704
+ attribute {
705
+ name: "to"
706
+ i: 7
707
+ type: INT
708
+ }
709
+ }
710
+ node {
711
+ output: "101"
712
+ name: "Constant_54"
713
+ op_type: "Constant"
714
+ attribute {
715
+ name: "value"
716
+ t {
717
+ data_type: 7
718
+ raw_data: "\002\000\000\000\000\000\000\000"
719
+ }
720
+ type: TENSOR
721
+ }
722
+ }
723
+ node {
724
+ input: "96"
725
+ input: "101"
726
+ output: "102"
727
+ name: "Mul_55"
728
+ op_type: "Mul"
729
+ }
730
+ node {
731
+ input: "90"
732
+ output: "104"
733
+ name: "Unsqueeze_56"
734
+ op_type: "Unsqueeze"
735
+ attribute {
736
+ name: "axes"
737
+ ints: 0
738
+ type: INTS
739
+ }
740
+ }
741
+ node {
742
+ input: "100"
743
+ output: "106"
744
+ name: "Unsqueeze_57"
745
+ op_type: "Unsqueeze"
746
+ attribute {
747
+ name: "axes"
748
+ ints: 0
749
+ type: INTS
750
+ }
751
+ }
752
+ node {
753
+ input: "96"
754
+ output: "107"
755
+ name: "Unsqueeze_58"
756
+ op_type: "Unsqueeze"
757
+ attribute {
758
+ name: "axes"
759
+ ints: 0
760
+ type: INTS
761
+ }
762
+ }
763
+ node {
764
+ input: "104"
765
+ input: "222"
766
+ input: "106"
767
+ input: "107"
768
+ output: "108"
769
+ name: "Concat_59"
770
+ op_type: "Concat"
771
+ attribute {
772
+ name: "axis"
773
+ i: 0
774
+ type: INT
775
+ }
776
+ }
777
+ node {
778
+ input: "87"
779
+ input: "108"
780
+ output: "109"
781
+ name: "Reshape_60"
782
+ op_type: "Reshape"
783
+ }
784
+ node {
785
+ input: "109"
786
+ output: "110"
787
+ name: "Transpose_61"
788
+ op_type: "Transpose"
789
+ attribute {
790
+ name: "perm"
791
+ ints: 0
792
+ ints: 2
793
+ ints: 3
794
+ ints: 1
795
+ type: INTS
796
+ }
797
+ }
798
+ node {
799
+ input: "90"
800
+ output: "111"
801
+ name: "Unsqueeze_62"
802
+ op_type: "Unsqueeze"
803
+ attribute {
804
+ name: "axes"
805
+ ints: 0
806
+ type: INTS
807
+ }
808
+ }
809
+ node {
810
+ input: "100"
811
+ output: "112"
812
+ name: "Unsqueeze_63"
813
+ op_type: "Unsqueeze"
814
+ attribute {
815
+ name: "axes"
816
+ ints: 0
817
+ type: INTS
818
+ }
819
+ }
820
+ node {
821
+ input: "102"
822
+ output: "113"
823
+ name: "Unsqueeze_64"
824
+ op_type: "Unsqueeze"
825
+ attribute {
826
+ name: "axes"
827
+ ints: 0
828
+ type: INTS
829
+ }
830
+ }
831
+ node {
832
+ input: "111"
833
+ input: "112"
834
+ input: "113"
835
+ output: "114"
836
+ name: "Concat_65"
837
+ op_type: "Concat"
838
+ attribute {
839
+ name: "axis"
840
+ i: 0
841
+ type: INT
842
+ }
843
+ }
844
+ node {
845
+ input: "110"
846
+ input: "114"
847
+ output: "115"
848
+ name: "Reshape_66"
849
+ op_type: "Reshape"
850
+ }
851
+ node {
852
+ input: "115"
853
+ input: "49"
854
+ output: "116"
855
+ name: "Concat_67"
856
+ op_type: "Concat"
857
+ attribute {
858
+ name: "axis"
859
+ i: 1
860
+ type: INT
861
+ }
862
+ }
863
+ node {
864
+ input: "116"
865
+ input: "up_net.1.0.weight"
866
+ input: "up_net.1.0.bias"
867
+ output: "117"
868
+ name: "Conv_68"
869
+ op_type: "Conv"
870
+ attribute {
871
+ name: "dilations"
872
+ ints: 1
873
+ type: INTS
874
+ }
875
+ attribute {
876
+ name: "group"
877
+ i: 1
878
+ type: INT
879
+ }
880
+ attribute {
881
+ name: "kernel_shape"
882
+ ints: 33
883
+ type: INTS
884
+ }
885
+ attribute {
886
+ name: "pads"
887
+ ints: 16
888
+ ints: 16
889
+ type: INTS
890
+ }
891
+ attribute {
892
+ name: "strides"
893
+ ints: 1
894
+ type: INTS
895
+ }
896
+ }
897
+ node {
898
+ input: "117"
899
+ output: "118"
900
+ name: "Relu_69"
901
+ op_type: "Relu"
902
+ }
903
+ node {
904
+ input: "118"
905
+ output: "119"
906
+ name: "Shape_70"
907
+ op_type: "Shape"
908
+ }
909
+ node {
910
+ output: "120"
911
+ name: "Constant_71"
912
+ op_type: "Constant"
913
+ attribute {
914
+ name: "value"
915
+ t {
916
+ data_type: 7
917
+ raw_data: "\000\000\000\000\000\000\000\000"
918
+ }
919
+ type: TENSOR
920
+ }
921
+ }
922
+ node {
923
+ input: "119"
924
+ input: "120"
925
+ output: "121"
926
+ name: "Gather_72"
927
+ op_type: "Gather"
928
+ attribute {
929
+ name: "axis"
930
+ i: 0
931
+ type: INT
932
+ }
933
+ }
934
+ node {
935
+ input: "118"
936
+ output: "122"
937
+ name: "Shape_73"
938
+ op_type: "Shape"
939
+ }
940
+ node {
941
+ output: "123"
942
+ name: "Constant_74"
943
+ op_type: "Constant"
944
+ attribute {
945
+ name: "value"
946
+ t {
947
+ data_type: 7
948
+ raw_data: "\001\000\000\000\000\000\000\000"
949
+ }
950
+ type: TENSOR
951
+ }
952
+ }
953
+ node {
954
+ input: "122"
955
+ input: "123"
956
+ output: "124"
957
+ name: "Gather_75"
958
+ op_type: "Gather"
959
+ attribute {
960
+ name: "axis"
961
+ i: 0
962
+ type: INT
963
+ }
964
+ }
965
+ node {
966
+ input: "118"
967
+ output: "125"
968
+ name: "Shape_76"
969
+ op_type: "Shape"
970
+ }
971
+ node {
972
+ output: "126"
973
+ name: "Constant_77"
974
+ op_type: "Constant"
975
+ attribute {
976
+ name: "value"
977
+ t {
978
+ data_type: 7
979
+ raw_data: "\002\000\000\000\000\000\000\000"
980
+ }
981
+ type: TENSOR
982
+ }
983
+ }
984
+ node {
985
+ input: "125"
986
+ input: "126"
987
+ output: "127"
988
+ name: "Gather_78"
989
+ op_type: "Gather"
990
+ attribute {
991
+ name: "axis"
992
+ i: 0
993
+ type: INT
994
+ }
995
+ }
996
+ node {
997
+ output: "128"
998
+ name: "Constant_79"
999
+ op_type: "Constant"
1000
+ attribute {
1001
+ name: "value"
1002
+ t {
1003
+ data_type: 7
1004
+ raw_data: "\002\000\000\000\000\000\000\000"
1005
+ }
1006
+ type: TENSOR
1007
+ }
1008
+ }
1009
+ node {
1010
+ input: "124"
1011
+ input: "128"
1012
+ output: "129"
1013
+ name: "Div_80"
1014
+ op_type: "Div"
1015
+ }
1016
+ node {
1017
+ input: "129"
1018
+ output: "130"
1019
+ name: "Cast_81"
1020
+ op_type: "Cast"
1021
+ attribute {
1022
+ name: "to"
1023
+ i: 7
1024
+ type: INT
1025
+ }
1026
+ }
1027
+ node {
1028
+ input: "130"
1029
+ output: "131"
1030
+ name: "Cast_82"
1031
+ op_type: "Cast"
1032
+ attribute {
1033
+ name: "to"
1034
+ i: 7
1035
+ type: INT
1036
+ }
1037
+ }
1038
+ node {
1039
+ output: "132"
1040
+ name: "Constant_83"
1041
+ op_type: "Constant"
1042
+ attribute {
1043
+ name: "value"
1044
+ t {
1045
+ data_type: 7
1046
+ raw_data: "\002\000\000\000\000\000\000\000"
1047
+ }
1048
+ type: TENSOR
1049
+ }
1050
+ }
1051
+ node {
1052
+ input: "127"
1053
+ input: "132"
1054
+ output: "133"
1055
+ name: "Mul_84"
1056
+ op_type: "Mul"
1057
+ }
1058
+ node {
1059
+ input: "121"
1060
+ output: "135"
1061
+ name: "Unsqueeze_85"
1062
+ op_type: "Unsqueeze"
1063
+ attribute {
1064
+ name: "axes"
1065
+ ints: 0
1066
+ type: INTS
1067
+ }
1068
+ }
1069
+ node {
1070
+ input: "131"
1071
+ output: "137"
1072
+ name: "Unsqueeze_86"
1073
+ op_type: "Unsqueeze"
1074
+ attribute {
1075
+ name: "axes"
1076
+ ints: 0
1077
+ type: INTS
1078
+ }
1079
+ }
1080
+ node {
1081
+ input: "127"
1082
+ output: "138"
1083
+ name: "Unsqueeze_87"
1084
+ op_type: "Unsqueeze"
1085
+ attribute {
1086
+ name: "axes"
1087
+ ints: 0
1088
+ type: INTS
1089
+ }
1090
+ }
1091
+ node {
1092
+ input: "135"
1093
+ input: "223"
1094
+ input: "137"
1095
+ input: "138"
1096
+ output: "139"
1097
+ name: "Concat_88"
1098
+ op_type: "Concat"
1099
+ attribute {
1100
+ name: "axis"
1101
+ i: 0
1102
+ type: INT
1103
+ }
1104
+ }
1105
+ node {
1106
+ input: "118"
1107
+ input: "139"
1108
+ output: "140"
1109
+ name: "Reshape_89"
1110
+ op_type: "Reshape"
1111
+ }
1112
+ node {
1113
+ input: "140"
1114
+ output: "141"
1115
+ name: "Transpose_90"
1116
+ op_type: "Transpose"
1117
+ attribute {
1118
+ name: "perm"
1119
+ ints: 0
1120
+ ints: 2
1121
+ ints: 3
1122
+ ints: 1
1123
+ type: INTS
1124
+ }
1125
+ }
1126
+ node {
1127
+ input: "121"
1128
+ output: "142"
1129
+ name: "Unsqueeze_91"
1130
+ op_type: "Unsqueeze"
1131
+ attribute {
1132
+ name: "axes"
1133
+ ints: 0
1134
+ type: INTS
1135
+ }
1136
+ }
1137
+ node {
1138
+ input: "131"
1139
+ output: "143"
1140
+ name: "Unsqueeze_92"
1141
+ op_type: "Unsqueeze"
1142
+ attribute {
1143
+ name: "axes"
1144
+ ints: 0
1145
+ type: INTS
1146
+ }
1147
+ }
1148
+ node {
1149
+ input: "133"
1150
+ output: "144"
1151
+ name: "Unsqueeze_93"
1152
+ op_type: "Unsqueeze"
1153
+ attribute {
1154
+ name: "axes"
1155
+ ints: 0
1156
+ type: INTS
1157
+ }
1158
+ }
1159
+ node {
1160
+ input: "142"
1161
+ input: "143"
1162
+ input: "144"
1163
+ output: "145"
1164
+ name: "Concat_94"
1165
+ op_type: "Concat"
1166
+ attribute {
1167
+ name: "axis"
1168
+ i: 0
1169
+ type: INT
1170
+ }
1171
+ }
1172
+ node {
1173
+ input: "141"
1174
+ input: "145"
1175
+ output: "146"
1176
+ name: "Reshape_95"
1177
+ op_type: "Reshape"
1178
+ }
1179
+ node {
1180
+ input: "146"
1181
+ input: "46"
1182
+ output: "147"
1183
+ name: "Concat_96"
1184
+ op_type: "Concat"
1185
+ attribute {
1186
+ name: "axis"
1187
+ i: 1
1188
+ type: INT
1189
+ }
1190
+ }
1191
+ node {
1192
+ input: "147"
1193
+ input: "up_net.2.0.weight"
1194
+ input: "up_net.2.0.bias"
1195
+ output: "148"
1196
+ name: "Conv_97"
1197
+ op_type: "Conv"
1198
+ attribute {
1199
+ name: "dilations"
1200
+ ints: 1
1201
+ type: INTS
1202
+ }
1203
+ attribute {
1204
+ name: "group"
1205
+ i: 1
1206
+ type: INT
1207
+ }
1208
+ attribute {
1209
+ name: "kernel_shape"
1210
+ ints: 65
1211
+ type: INTS
1212
+ }
1213
+ attribute {
1214
+ name: "pads"
1215
+ ints: 32
1216
+ ints: 32
1217
+ type: INTS
1218
+ }
1219
+ attribute {
1220
+ name: "strides"
1221
+ ints: 1
1222
+ type: INTS
1223
+ }
1224
+ }
1225
+ node {
1226
+ input: "148"
1227
+ output: "149"
1228
+ name: "Relu_98"
1229
+ op_type: "Relu"
1230
+ }
1231
+ node {
1232
+ input: "149"
1233
+ output: "150"
1234
+ name: "Shape_99"
1235
+ op_type: "Shape"
1236
+ }
1237
+ node {
1238
+ output: "151"
1239
+ name: "Constant_100"
1240
+ op_type: "Constant"
1241
+ attribute {
1242
+ name: "value"
1243
+ t {
1244
+ data_type: 7
1245
+ raw_data: "\000\000\000\000\000\000\000\000"
1246
+ }
1247
+ type: TENSOR
1248
+ }
1249
+ }
1250
+ node {
1251
+ input: "150"
1252
+ input: "151"
1253
+ output: "152"
1254
+ name: "Gather_101"
1255
+ op_type: "Gather"
1256
+ attribute {
1257
+ name: "axis"
1258
+ i: 0
1259
+ type: INT
1260
+ }
1261
+ }
1262
+ node {
1263
+ input: "149"
1264
+ output: "153"
1265
+ name: "Shape_102"
1266
+ op_type: "Shape"
1267
+ }
1268
+ node {
1269
+ output: "154"
1270
+ name: "Constant_103"
1271
+ op_type: "Constant"
1272
+ attribute {
1273
+ name: "value"
1274
+ t {
1275
+ data_type: 7
1276
+ raw_data: "\001\000\000\000\000\000\000\000"
1277
+ }
1278
+ type: TENSOR
1279
+ }
1280
+ }
1281
+ node {
1282
+ input: "153"
1283
+ input: "154"
1284
+ output: "155"
1285
+ name: "Gather_104"
1286
+ op_type: "Gather"
1287
+ attribute {
1288
+ name: "axis"
1289
+ i: 0
1290
+ type: INT
1291
+ }
1292
+ }
1293
+ node {
1294
+ input: "149"
1295
+ output: "156"
1296
+ name: "Shape_105"
1297
+ op_type: "Shape"
1298
+ }
1299
+ node {
1300
+ output: "157"
1301
+ name: "Constant_106"
1302
+ op_type: "Constant"
1303
+ attribute {
1304
+ name: "value"
1305
+ t {
1306
+ data_type: 7
1307
+ raw_data: "\002\000\000\000\000\000\000\000"
1308
+ }
1309
+ type: TENSOR
1310
+ }
1311
+ }
1312
+ node {
1313
+ input: "156"
1314
+ input: "157"
1315
+ output: "158"
1316
+ name: "Gather_107"
1317
+ op_type: "Gather"
1318
+ attribute {
1319
+ name: "axis"
1320
+ i: 0
1321
+ type: INT
1322
+ }
1323
+ }
1324
+ node {
1325
+ output: "159"
1326
+ name: "Constant_108"
1327
+ op_type: "Constant"
1328
+ attribute {
1329
+ name: "value"
1330
+ t {
1331
+ data_type: 7
1332
+ raw_data: "\002\000\000\000\000\000\000\000"
1333
+ }
1334
+ type: TENSOR
1335
+ }
1336
+ }
1337
+ node {
1338
+ input: "155"
1339
+ input: "159"
1340
+ output: "160"
1341
+ name: "Div_109"
1342
+ op_type: "Div"
1343
+ }
1344
+ node {
1345
+ input: "160"
1346
+ output: "161"
1347
+ name: "Cast_110"
1348
+ op_type: "Cast"
1349
+ attribute {
1350
+ name: "to"
1351
+ i: 7
1352
+ type: INT
1353
+ }
1354
+ }
1355
+ node {
1356
+ input: "161"
1357
+ output: "162"
1358
+ name: "Cast_111"
1359
+ op_type: "Cast"
1360
+ attribute {
1361
+ name: "to"
1362
+ i: 7
1363
+ type: INT
1364
+ }
1365
+ }
1366
+ node {
1367
+ output: "163"
1368
+ name: "Constant_112"
1369
+ op_type: "Constant"
1370
+ attribute {
1371
+ name: "value"
1372
+ t {
1373
+ data_type: 7
1374
+ raw_data: "\002\000\000\000\000\000\000\000"
1375
+ }
1376
+ type: TENSOR
1377
+ }
1378
+ }
1379
+ node {
1380
+ input: "158"
1381
+ input: "163"
1382
+ output: "164"
1383
+ name: "Mul_113"
1384
+ op_type: "Mul"
1385
+ }
1386
+ node {
1387
+ input: "152"
1388
+ output: "166"
1389
+ name: "Unsqueeze_114"
1390
+ op_type: "Unsqueeze"
1391
+ attribute {
1392
+ name: "axes"
1393
+ ints: 0
1394
+ type: INTS
1395
+ }
1396
+ }
1397
+ node {
1398
+ input: "162"
1399
+ output: "168"
1400
+ name: "Unsqueeze_115"
1401
+ op_type: "Unsqueeze"
1402
+ attribute {
1403
+ name: "axes"
1404
+ ints: 0
1405
+ type: INTS
1406
+ }
1407
+ }
1408
+ node {
1409
+ input: "158"
1410
+ output: "169"
1411
+ name: "Unsqueeze_116"
1412
+ op_type: "Unsqueeze"
1413
+ attribute {
1414
+ name: "axes"
1415
+ ints: 0
1416
+ type: INTS
1417
+ }
1418
+ }
1419
+ node {
1420
+ input: "166"
1421
+ input: "224"
1422
+ input: "168"
1423
+ input: "169"
1424
+ output: "170"
1425
+ name: "Concat_117"
1426
+ op_type: "Concat"
1427
+ attribute {
1428
+ name: "axis"
1429
+ i: 0
1430
+ type: INT
1431
+ }
1432
+ }
1433
+ node {
1434
+ input: "149"
1435
+ input: "170"
1436
+ output: "171"
1437
+ name: "Reshape_118"
1438
+ op_type: "Reshape"
1439
+ }
1440
+ node {
1441
+ input: "171"
1442
+ output: "172"
1443
+ name: "Transpose_119"
1444
+ op_type: "Transpose"
1445
+ attribute {
1446
+ name: "perm"
1447
+ ints: 0
1448
+ ints: 2
1449
+ ints: 3
1450
+ ints: 1
1451
+ type: INTS
1452
+ }
1453
+ }
1454
+ node {
1455
+ input: "152"
1456
+ output: "173"
1457
+ name: "Unsqueeze_120"
1458
+ op_type: "Unsqueeze"
1459
+ attribute {
1460
+ name: "axes"
1461
+ ints: 0
1462
+ type: INTS
1463
+ }
1464
+ }
1465
+ node {
1466
+ input: "162"
1467
+ output: "174"
1468
+ name: "Unsqueeze_121"
1469
+ op_type: "Unsqueeze"
1470
+ attribute {
1471
+ name: "axes"
1472
+ ints: 0
1473
+ type: INTS
1474
+ }
1475
+ }
1476
+ node {
1477
+ input: "164"
1478
+ output: "175"
1479
+ name: "Unsqueeze_122"
1480
+ op_type: "Unsqueeze"
1481
+ attribute {
1482
+ name: "axes"
1483
+ ints: 0
1484
+ type: INTS
1485
+ }
1486
+ }
1487
+ node {
1488
+ input: "173"
1489
+ input: "174"
1490
+ input: "175"
1491
+ output: "176"
1492
+ name: "Concat_123"
1493
+ op_type: "Concat"
1494
+ attribute {
1495
+ name: "axis"
1496
+ i: 0
1497
+ type: INT
1498
+ }
1499
+ }
1500
+ node {
1501
+ input: "172"
1502
+ input: "176"
1503
+ output: "177"
1504
+ name: "Reshape_124"
1505
+ op_type: "Reshape"
1506
+ }
1507
+ node {
1508
+ input: "177"
1509
+ input: "43"
1510
+ output: "178"
1511
+ name: "Concat_125"
1512
+ op_type: "Concat"
1513
+ attribute {
1514
+ name: "axis"
1515
+ i: 1
1516
+ type: INT
1517
+ }
1518
+ }
1519
+ node {
1520
+ input: "178"
1521
+ input: "up_net.3.0.weight"
1522
+ input: "up_net.3.0.bias"
1523
+ output: "179"
1524
+ name: "Conv_126"
1525
+ op_type: "Conv"
1526
+ attribute {
1527
+ name: "dilations"
1528
+ ints: 1
1529
+ type: INTS
1530
+ }
1531
+ attribute {
1532
+ name: "group"
1533
+ i: 1
1534
+ type: INT
1535
+ }
1536
+ attribute {
1537
+ name: "kernel_shape"
1538
+ ints: 9
1539
+ type: INTS
1540
+ }
1541
+ attribute {
1542
+ name: "pads"
1543
+ ints: 4
1544
+ ints: 4
1545
+ type: INTS
1546
+ }
1547
+ attribute {
1548
+ name: "strides"
1549
+ ints: 1
1550
+ type: INTS
1551
+ }
1552
+ }
1553
+ node {
1554
+ input: "179"
1555
+ output: "180"
1556
+ name: "Shape_127"
1557
+ op_type: "Shape"
1558
+ }
1559
+ node {
1560
+ output: "181"
1561
+ name: "Constant_128"
1562
+ op_type: "Constant"
1563
+ attribute {
1564
+ name: "value"
1565
+ t {
1566
+ data_type: 7
1567
+ raw_data: "\000\000\000\000\000\000\000\000"
1568
+ }
1569
+ type: TENSOR
1570
+ }
1571
+ }
1572
+ node {
1573
+ input: "180"
1574
+ input: "181"
1575
+ output: "182"
1576
+ name: "Gather_129"
1577
+ op_type: "Gather"
1578
+ attribute {
1579
+ name: "axis"
1580
+ i: 0
1581
+ type: INT
1582
+ }
1583
+ }
1584
+ node {
1585
+ input: "179"
1586
+ output: "183"
1587
+ name: "Shape_130"
1588
+ op_type: "Shape"
1589
+ }
1590
+ node {
1591
+ output: "184"
1592
+ name: "Constant_131"
1593
+ op_type: "Constant"
1594
+ attribute {
1595
+ name: "value"
1596
+ t {
1597
+ data_type: 7
1598
+ raw_data: "\001\000\000\000\000\000\000\000"
1599
+ }
1600
+ type: TENSOR
1601
+ }
1602
+ }
1603
+ node {
1604
+ input: "183"
1605
+ input: "184"
1606
+ output: "185"
1607
+ name: "Gather_132"
1608
+ op_type: "Gather"
1609
+ attribute {
1610
+ name: "axis"
1611
+ i: 0
1612
+ type: INT
1613
+ }
1614
+ }
1615
+ node {
1616
+ input: "179"
1617
+ output: "186"
1618
+ name: "Shape_133"
1619
+ op_type: "Shape"
1620
+ }
1621
+ node {
1622
+ output: "187"
1623
+ name: "Constant_134"
1624
+ op_type: "Constant"
1625
+ attribute {
1626
+ name: "value"
1627
+ t {
1628
+ data_type: 7
1629
+ raw_data: "\002\000\000\000\000\000\000\000"
1630
+ }
1631
+ type: TENSOR
1632
+ }
1633
+ }
1634
+ node {
1635
+ input: "186"
1636
+ input: "187"
1637
+ output: "188"
1638
+ name: "Gather_135"
1639
+ op_type: "Gather"
1640
+ attribute {
1641
+ name: "axis"
1642
+ i: 0
1643
+ type: INT
1644
+ }
1645
+ }
1646
+ node {
1647
+ output: "189"
1648
+ name: "Constant_136"
1649
+ op_type: "Constant"
1650
+ attribute {
1651
+ name: "value"
1652
+ t {
1653
+ data_type: 7
1654
+ raw_data: "\002\000\000\000\000\000\000\000"
1655
+ }
1656
+ type: TENSOR
1657
+ }
1658
+ }
1659
+ node {
1660
+ input: "185"
1661
+ input: "189"
1662
+ output: "190"
1663
+ name: "Div_137"
1664
+ op_type: "Div"
1665
+ }
1666
+ node {
1667
+ input: "190"
1668
+ output: "191"
1669
+ name: "Cast_138"
1670
+ op_type: "Cast"
1671
+ attribute {
1672
+ name: "to"
1673
+ i: 7
1674
+ type: INT
1675
+ }
1676
+ }
1677
+ node {
1678
+ input: "191"
1679
+ output: "192"
1680
+ name: "Cast_139"
1681
+ op_type: "Cast"
1682
+ attribute {
1683
+ name: "to"
1684
+ i: 7
1685
+ type: INT
1686
+ }
1687
+ }
1688
+ node {
1689
+ output: "193"
1690
+ name: "Constant_140"
1691
+ op_type: "Constant"
1692
+ attribute {
1693
+ name: "value"
1694
+ t {
1695
+ data_type: 7
1696
+ raw_data: "\002\000\000\000\000\000\000\000"
1697
+ }
1698
+ type: TENSOR
1699
+ }
1700
+ }
1701
+ node {
1702
+ input: "188"
1703
+ input: "193"
1704
+ output: "194"
1705
+ name: "Mul_141"
1706
+ op_type: "Mul"
1707
+ }
1708
+ node {
1709
+ input: "182"
1710
+ output: "196"
1711
+ name: "Unsqueeze_142"
1712
+ op_type: "Unsqueeze"
1713
+ attribute {
1714
+ name: "axes"
1715
+ ints: 0
1716
+ type: INTS
1717
+ }
1718
+ }
1719
+ node {
1720
+ input: "192"
1721
+ output: "198"
1722
+ name: "Unsqueeze_143"
1723
+ op_type: "Unsqueeze"
1724
+ attribute {
1725
+ name: "axes"
1726
+ ints: 0
1727
+ type: INTS
1728
+ }
1729
+ }
1730
+ node {
1731
+ input: "188"
1732
+ output: "199"
1733
+ name: "Unsqueeze_144"
1734
+ op_type: "Unsqueeze"
1735
+ attribute {
1736
+ name: "axes"
1737
+ ints: 0
1738
+ type: INTS
1739
+ }
1740
+ }
1741
+ node {
1742
+ input: "196"
1743
+ input: "225"
1744
+ input: "198"
1745
+ input: "199"
1746
+ output: "200"
1747
+ name: "Concat_145"
1748
+ op_type: "Concat"
1749
+ attribute {
1750
+ name: "axis"
1751
+ i: 0
1752
+ type: INT
1753
+ }
1754
+ }
1755
+ node {
1756
+ input: "179"
1757
+ input: "200"
1758
+ output: "201"
1759
+ name: "Reshape_146"
1760
+ op_type: "Reshape"
1761
+ }
1762
+ node {
1763
+ input: "201"
1764
+ output: "202"
1765
+ name: "Transpose_147"
1766
+ op_type: "Transpose"
1767
+ attribute {
1768
+ name: "perm"
1769
+ ints: 0
1770
+ ints: 2
1771
+ ints: 3
1772
+ ints: 1
1773
+ type: INTS
1774
+ }
1775
+ }
1776
+ node {
1777
+ input: "182"
1778
+ output: "203"
1779
+ name: "Unsqueeze_148"
1780
+ op_type: "Unsqueeze"
1781
+ attribute {
1782
+ name: "axes"
1783
+ ints: 0
1784
+ type: INTS
1785
+ }
1786
+ }
1787
+ node {
1788
+ input: "192"
1789
+ output: "204"
1790
+ name: "Unsqueeze_149"
1791
+ op_type: "Unsqueeze"
1792
+ attribute {
1793
+ name: "axes"
1794
+ ints: 0
1795
+ type: INTS
1796
+ }
1797
+ }
1798
+ node {
1799
+ input: "194"
1800
+ output: "205"
1801
+ name: "Unsqueeze_150"
1802
+ op_type: "Unsqueeze"
1803
+ attribute {
1804
+ name: "axes"
1805
+ ints: 0
1806
+ type: INTS
1807
+ }
1808
+ }
1809
+ node {
1810
+ input: "203"
1811
+ input: "204"
1812
+ input: "205"
1813
+ output: "206"
1814
+ name: "Concat_151"
1815
+ op_type: "Concat"
1816
+ attribute {
1817
+ name: "axis"
1818
+ i: 0
1819
+ type: INT
1820
+ }
1821
+ }
1822
+ node {
1823
+ input: "202"
1824
+ input: "206"
1825
+ output: "207"
1826
+ name: "Reshape_152"
1827
+ op_type: "Reshape"
1828
+ }
1829
+ node {
1830
+ input: "207"
1831
+ input: "x"
1832
+ output: "y"
1833
+ name: "Add_153"
1834
+ op_type: "Add"
1835
+ }
1836
+ initializer {
1837
+ dims: 128
1838
+ dims: 2
1839
+ dims: 65
1840
+ data_type: 1
1841
+ name: "210"
1842
+ }
1843
+ initializer {
1844
+ dims: 128
1845
+ data_type: 1
1846
+ name: "211"
1847
+ }
1848
+ initializer {
1849
+ dims: 256
1850
+ dims: 128
1851
+ dims: 33
1852
+ data_type: 1
1853
+ name: "213"
1854
+ }
1855
+ initializer {
1856
+ dims: 256
1857
+ data_type: 1
1858
+ name: "214"
1859
+ }
1860
+ initializer {
1861
+ dims: 512
1862
+ dims: 256
1863
+ dims: 17
1864
+ data_type: 1
1865
+ name: "216"
1866
+ }
1867
+ initializer {
1868
+ dims: 512
1869
+ data_type: 1
1870
+ name: "217"
1871
+ }
1872
+ initializer {
1873
+ dims: 512
1874
+ dims: 512
1875
+ dims: 9
1876
+ data_type: 1
1877
+ name: "219"
1878
+ }
1879
+ initializer {
1880
+ dims: 512
1881
+ data_type: 1
1882
+ name: "220"
1883
+ }
1884
+ initializer {
1885
+ dims: 1
1886
+ data_type: 7
1887
+ name: "221"
1888
+ raw_data: "\002\000\000\000\000\000\000\000"
1889
+ }
1890
+ initializer {
1891
+ dims: 1
1892
+ data_type: 7
1893
+ name: "222"
1894
+ raw_data: "\002\000\000\000\000\000\000\000"
1895
+ }
1896
+ initializer {
1897
+ dims: 1
1898
+ data_type: 7
1899
+ name: "223"
1900
+ raw_data: "\002\000\000\000\000\000\000\000"
1901
+ }
1902
+ initializer {
1903
+ dims: 1
1904
+ data_type: 7
1905
+ name: "224"
1906
+ raw_data: "\002\000\000\000\000\000\000\000"
1907
+ }
1908
+ initializer {
1909
+ dims: 1
1910
+ data_type: 7
1911
+ name: "225"
1912
+ raw_data: "\002\000\000\000\000\000\000\000"
1913
+ }
1914
+ initializer {
1915
+ dims: 512
1916
+ data_type: 1
1917
+ name: "bottleneck.0.bias"
1918
+ }
1919
+ initializer {
1920
+ dims: 512
1921
+ dims: 512
1922
+ dims: 9
1923
+ data_type: 1
1924
+ name: "bottleneck.0.weight"
1925
+ }
1926
+ initializer {
1927
+ dims: 1024
1928
+ data_type: 1
1929
+ name: "bottleneck.2.bias"
1930
+ }
1931
+ initializer {
1932
+ dims: 1024
1933
+ dims: 512
1934
+ dims: 9
1935
+ data_type: 1
1936
+ name: "bottleneck.2.weight"
1937
+ }
1938
+ initializer {
1939
+ dims: 1024
1940
+ data_type: 1
1941
+ name: "up_net.0.0.bias"
1942
+ }
1943
+ initializer {
1944
+ dims: 1024
1945
+ dims: 1024
1946
+ dims: 17
1947
+ data_type: 1
1948
+ name: "up_net.0.0.weight"
1949
+ }
1950
+ initializer {
1951
+ dims: 512
1952
+ data_type: 1
1953
+ name: "up_net.1.0.bias"
1954
+ }
1955
+ initializer {
1956
+ dims: 512
1957
+ dims: 1024
1958
+ dims: 33
1959
+ data_type: 1
1960
+ name: "up_net.1.0.weight"
1961
+ }
1962
+ initializer {
1963
+ dims: 256
1964
+ data_type: 1
1965
+ name: "up_net.2.0.bias"
1966
+ }
1967
+ initializer {
1968
+ dims: 256
1969
+ dims: 512
1970
+ dims: 65
1971
+ data_type: 1
1972
+ name: "up_net.2.0.weight"
1973
+ }
1974
+ initializer {
1975
+ dims: 4
1976
+ data_type: 1
1977
+ name: "up_net.3.0.bias"
1978
+ raw_data: "[\005R\2729\022\002\273\222 \236\272\307\340\210\272"
1979
+ }
1980
+ initializer {
1981
+ dims: 4
1982
+ dims: 256
1983
+ dims: 9
1984
+ data_type: 1
1985
+ name: "up_net.3.0.weight"
1986
+ }
1987
+ input {
1988
+ name: "x"
1989
+ type {
1990
+ tensor_type {
1991
+ elem_type: 1
1992
+ shape {
1993
+ dim {
1994
+ dim_param: "batch_size"
1995
+ }
1996
+ dim {
1997
+ dim_value: 2
1998
+ }
1999
+ dim {
2000
+ dim_value: 8192
2001
+ }
2002
+ }
2003
+ }
2004
+ }
2005
+ }
2006
+ output {
2007
+ name: "y"
2008
+ type {
2009
+ tensor_type {
2010
+ elem_type: 1
2011
+ shape {
2012
+ dim {
2013
+ dim_param: "batch_size"
2014
+ }
2015
+ dim {
2016
+ dim_value: 2
2017
+ }
2018
+ dim {
2019
+ dim_value: 8192
2020
+ }
2021
+ }
2022
+ }
2023
+ }
2024
+ }
2025
+ }
2026
+ opset_import {
2027
+ version: 11
2028
+ }
models/ailia-models/unetda.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a3ea3579de92dd42313b96d11a8e8d07ee43b1ba92e2cb20f3500f0e279220e
3
+ size 225706017
models/ailia-models/unetda.onnx.prototxt ADDED
@@ -0,0 +1,2028 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "down_net.0.0.weight"
10
+ input: "down_net.0.0.bias"
11
+ output: "21"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 65
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 32
32
+ ints: 32
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 2
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "21"
43
+ output: "22"
44
+ name: "Relu_1"
45
+ op_type: "Relu"
46
+ }
47
+ node {
48
+ input: "22"
49
+ input: "down_net.1.0.weight"
50
+ input: "down_net.1.0.bias"
51
+ output: "23"
52
+ name: "Conv_2"
53
+ op_type: "Conv"
54
+ attribute {
55
+ name: "dilations"
56
+ ints: 1
57
+ type: INTS
58
+ }
59
+ attribute {
60
+ name: "group"
61
+ i: 1
62
+ type: INT
63
+ }
64
+ attribute {
65
+ name: "kernel_shape"
66
+ ints: 33
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "pads"
71
+ ints: 16
72
+ ints: 16
73
+ type: INTS
74
+ }
75
+ attribute {
76
+ name: "strides"
77
+ ints: 2
78
+ type: INTS
79
+ }
80
+ }
81
+ node {
82
+ input: "23"
83
+ output: "24"
84
+ name: "Relu_3"
85
+ op_type: "Relu"
86
+ }
87
+ node {
88
+ input: "24"
89
+ input: "down_net.2.0.weight"
90
+ input: "down_net.2.0.bias"
91
+ output: "25"
92
+ name: "Conv_4"
93
+ op_type: "Conv"
94
+ attribute {
95
+ name: "dilations"
96
+ ints: 1
97
+ type: INTS
98
+ }
99
+ attribute {
100
+ name: "group"
101
+ i: 1
102
+ type: INT
103
+ }
104
+ attribute {
105
+ name: "kernel_shape"
106
+ ints: 17
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "pads"
111
+ ints: 8
112
+ ints: 8
113
+ type: INTS
114
+ }
115
+ attribute {
116
+ name: "strides"
117
+ ints: 2
118
+ type: INTS
119
+ }
120
+ }
121
+ node {
122
+ input: "25"
123
+ output: "26"
124
+ name: "Relu_5"
125
+ op_type: "Relu"
126
+ }
127
+ node {
128
+ input: "26"
129
+ input: "down_net.3.0.weight"
130
+ input: "down_net.3.0.bias"
131
+ output: "27"
132
+ name: "Conv_6"
133
+ op_type: "Conv"
134
+ attribute {
135
+ name: "dilations"
136
+ ints: 1
137
+ type: INTS
138
+ }
139
+ attribute {
140
+ name: "group"
141
+ i: 1
142
+ type: INT
143
+ }
144
+ attribute {
145
+ name: "kernel_shape"
146
+ ints: 9
147
+ type: INTS
148
+ }
149
+ attribute {
150
+ name: "pads"
151
+ ints: 4
152
+ ints: 4
153
+ type: INTS
154
+ }
155
+ attribute {
156
+ name: "strides"
157
+ ints: 2
158
+ type: INTS
159
+ }
160
+ }
161
+ node {
162
+ input: "27"
163
+ output: "28"
164
+ name: "Relu_7"
165
+ op_type: "Relu"
166
+ }
167
+ node {
168
+ input: "28"
169
+ input: "bottleneck.0.weight"
170
+ input: "bottleneck.0.bias"
171
+ output: "29"
172
+ name: "Conv_8"
173
+ op_type: "Conv"
174
+ attribute {
175
+ name: "dilations"
176
+ ints: 1
177
+ type: INTS
178
+ }
179
+ attribute {
180
+ name: "group"
181
+ i: 1
182
+ type: INT
183
+ }
184
+ attribute {
185
+ name: "kernel_shape"
186
+ ints: 9
187
+ type: INTS
188
+ }
189
+ attribute {
190
+ name: "pads"
191
+ ints: 4
192
+ ints: 4
193
+ type: INTS
194
+ }
195
+ attribute {
196
+ name: "strides"
197
+ ints: 2
198
+ type: INTS
199
+ }
200
+ }
201
+ node {
202
+ input: "29"
203
+ output: "30"
204
+ name: "Relu_9"
205
+ op_type: "Relu"
206
+ }
207
+ node {
208
+ input: "30"
209
+ input: "bottleneck.2.weight"
210
+ input: "bottleneck.2.bias"
211
+ output: "31"
212
+ name: "Conv_10"
213
+ op_type: "Conv"
214
+ attribute {
215
+ name: "dilations"
216
+ ints: 1
217
+ type: INTS
218
+ }
219
+ attribute {
220
+ name: "group"
221
+ i: 1
222
+ type: INT
223
+ }
224
+ attribute {
225
+ name: "kernel_shape"
226
+ ints: 9
227
+ type: INTS
228
+ }
229
+ attribute {
230
+ name: "pads"
231
+ ints: 4
232
+ ints: 4
233
+ type: INTS
234
+ }
235
+ attribute {
236
+ name: "strides"
237
+ ints: 1
238
+ type: INTS
239
+ }
240
+ }
241
+ node {
242
+ input: "31"
243
+ output: "32"
244
+ name: "Relu_11"
245
+ op_type: "Relu"
246
+ }
247
+ node {
248
+ input: "32"
249
+ output: "33"
250
+ name: "Shape_12"
251
+ op_type: "Shape"
252
+ }
253
+ node {
254
+ output: "34"
255
+ name: "Constant_13"
256
+ op_type: "Constant"
257
+ attribute {
258
+ name: "value"
259
+ t {
260
+ data_type: 7
261
+ raw_data: "\000\000\000\000\000\000\000\000"
262
+ }
263
+ type: TENSOR
264
+ }
265
+ }
266
+ node {
267
+ input: "33"
268
+ input: "34"
269
+ output: "35"
270
+ name: "Gather_14"
271
+ op_type: "Gather"
272
+ attribute {
273
+ name: "axis"
274
+ i: 0
275
+ type: INT
276
+ }
277
+ }
278
+ node {
279
+ input: "32"
280
+ output: "36"
281
+ name: "Shape_15"
282
+ op_type: "Shape"
283
+ }
284
+ node {
285
+ output: "37"
286
+ name: "Constant_16"
287
+ op_type: "Constant"
288
+ attribute {
289
+ name: "value"
290
+ t {
291
+ data_type: 7
292
+ raw_data: "\001\000\000\000\000\000\000\000"
293
+ }
294
+ type: TENSOR
295
+ }
296
+ }
297
+ node {
298
+ input: "36"
299
+ input: "37"
300
+ output: "38"
301
+ name: "Gather_17"
302
+ op_type: "Gather"
303
+ attribute {
304
+ name: "axis"
305
+ i: 0
306
+ type: INT
307
+ }
308
+ }
309
+ node {
310
+ input: "32"
311
+ output: "39"
312
+ name: "Shape_18"
313
+ op_type: "Shape"
314
+ }
315
+ node {
316
+ output: "40"
317
+ name: "Constant_19"
318
+ op_type: "Constant"
319
+ attribute {
320
+ name: "value"
321
+ t {
322
+ data_type: 7
323
+ raw_data: "\002\000\000\000\000\000\000\000"
324
+ }
325
+ type: TENSOR
326
+ }
327
+ }
328
+ node {
329
+ input: "39"
330
+ input: "40"
331
+ output: "41"
332
+ name: "Gather_20"
333
+ op_type: "Gather"
334
+ attribute {
335
+ name: "axis"
336
+ i: 0
337
+ type: INT
338
+ }
339
+ }
340
+ node {
341
+ output: "42"
342
+ name: "Constant_21"
343
+ op_type: "Constant"
344
+ attribute {
345
+ name: "value"
346
+ t {
347
+ data_type: 7
348
+ raw_data: "\002\000\000\000\000\000\000\000"
349
+ }
350
+ type: TENSOR
351
+ }
352
+ }
353
+ node {
354
+ input: "38"
355
+ input: "42"
356
+ output: "43"
357
+ name: "Div_22"
358
+ op_type: "Div"
359
+ }
360
+ node {
361
+ input: "43"
362
+ output: "44"
363
+ name: "Cast_23"
364
+ op_type: "Cast"
365
+ attribute {
366
+ name: "to"
367
+ i: 7
368
+ type: INT
369
+ }
370
+ }
371
+ node {
372
+ input: "44"
373
+ output: "45"
374
+ name: "Cast_24"
375
+ op_type: "Cast"
376
+ attribute {
377
+ name: "to"
378
+ i: 7
379
+ type: INT
380
+ }
381
+ }
382
+ node {
383
+ output: "46"
384
+ name: "Constant_25"
385
+ op_type: "Constant"
386
+ attribute {
387
+ name: "value"
388
+ t {
389
+ data_type: 7
390
+ raw_data: "\002\000\000\000\000\000\000\000"
391
+ }
392
+ type: TENSOR
393
+ }
394
+ }
395
+ node {
396
+ input: "41"
397
+ input: "46"
398
+ output: "47"
399
+ name: "Mul_26"
400
+ op_type: "Mul"
401
+ }
402
+ node {
403
+ input: "35"
404
+ output: "49"
405
+ name: "Unsqueeze_27"
406
+ op_type: "Unsqueeze"
407
+ attribute {
408
+ name: "axes"
409
+ ints: 0
410
+ type: INTS
411
+ }
412
+ }
413
+ node {
414
+ input: "45"
415
+ output: "51"
416
+ name: "Unsqueeze_28"
417
+ op_type: "Unsqueeze"
418
+ attribute {
419
+ name: "axes"
420
+ ints: 0
421
+ type: INTS
422
+ }
423
+ }
424
+ node {
425
+ input: "41"
426
+ output: "52"
427
+ name: "Unsqueeze_29"
428
+ op_type: "Unsqueeze"
429
+ attribute {
430
+ name: "axes"
431
+ ints: 0
432
+ type: INTS
433
+ }
434
+ }
435
+ node {
436
+ input: "49"
437
+ input: "185"
438
+ input: "51"
439
+ input: "52"
440
+ output: "53"
441
+ name: "Concat_30"
442
+ op_type: "Concat"
443
+ attribute {
444
+ name: "axis"
445
+ i: 0
446
+ type: INT
447
+ }
448
+ }
449
+ node {
450
+ input: "32"
451
+ input: "53"
452
+ output: "54"
453
+ name: "Reshape_31"
454
+ op_type: "Reshape"
455
+ }
456
+ node {
457
+ input: "54"
458
+ output: "55"
459
+ name: "Transpose_32"
460
+ op_type: "Transpose"
461
+ attribute {
462
+ name: "perm"
463
+ ints: 0
464
+ ints: 2
465
+ ints: 3
466
+ ints: 1
467
+ type: INTS
468
+ }
469
+ }
470
+ node {
471
+ input: "35"
472
+ output: "56"
473
+ name: "Unsqueeze_33"
474
+ op_type: "Unsqueeze"
475
+ attribute {
476
+ name: "axes"
477
+ ints: 0
478
+ type: INTS
479
+ }
480
+ }
481
+ node {
482
+ input: "45"
483
+ output: "57"
484
+ name: "Unsqueeze_34"
485
+ op_type: "Unsqueeze"
486
+ attribute {
487
+ name: "axes"
488
+ ints: 0
489
+ type: INTS
490
+ }
491
+ }
492
+ node {
493
+ input: "47"
494
+ output: "58"
495
+ name: "Unsqueeze_35"
496
+ op_type: "Unsqueeze"
497
+ attribute {
498
+ name: "axes"
499
+ ints: 0
500
+ type: INTS
501
+ }
502
+ }
503
+ node {
504
+ input: "56"
505
+ input: "57"
506
+ input: "58"
507
+ output: "59"
508
+ name: "Concat_36"
509
+ op_type: "Concat"
510
+ attribute {
511
+ name: "axis"
512
+ i: 0
513
+ type: INT
514
+ }
515
+ }
516
+ node {
517
+ input: "55"
518
+ input: "59"
519
+ output: "60"
520
+ name: "Reshape_37"
521
+ op_type: "Reshape"
522
+ }
523
+ node {
524
+ input: "60"
525
+ input: "28"
526
+ output: "61"
527
+ name: "Concat_38"
528
+ op_type: "Concat"
529
+ attribute {
530
+ name: "axis"
531
+ i: 1
532
+ type: INT
533
+ }
534
+ }
535
+ node {
536
+ input: "61"
537
+ input: "up_net.0.0.weight"
538
+ input: "up_net.0.0.bias"
539
+ output: "62"
540
+ name: "Conv_39"
541
+ op_type: "Conv"
542
+ attribute {
543
+ name: "dilations"
544
+ ints: 1
545
+ type: INTS
546
+ }
547
+ attribute {
548
+ name: "group"
549
+ i: 1
550
+ type: INT
551
+ }
552
+ attribute {
553
+ name: "kernel_shape"
554
+ ints: 17
555
+ type: INTS
556
+ }
557
+ attribute {
558
+ name: "pads"
559
+ ints: 8
560
+ ints: 8
561
+ type: INTS
562
+ }
563
+ attribute {
564
+ name: "strides"
565
+ ints: 1
566
+ type: INTS
567
+ }
568
+ }
569
+ node {
570
+ input: "62"
571
+ output: "63"
572
+ name: "Relu_40"
573
+ op_type: "Relu"
574
+ }
575
+ node {
576
+ input: "63"
577
+ output: "64"
578
+ name: "Shape_41"
579
+ op_type: "Shape"
580
+ }
581
+ node {
582
+ output: "65"
583
+ name: "Constant_42"
584
+ op_type: "Constant"
585
+ attribute {
586
+ name: "value"
587
+ t {
588
+ data_type: 7
589
+ raw_data: "\000\000\000\000\000\000\000\000"
590
+ }
591
+ type: TENSOR
592
+ }
593
+ }
594
+ node {
595
+ input: "64"
596
+ input: "65"
597
+ output: "66"
598
+ name: "Gather_43"
599
+ op_type: "Gather"
600
+ attribute {
601
+ name: "axis"
602
+ i: 0
603
+ type: INT
604
+ }
605
+ }
606
+ node {
607
+ input: "63"
608
+ output: "67"
609
+ name: "Shape_44"
610
+ op_type: "Shape"
611
+ }
612
+ node {
613
+ output: "68"
614
+ name: "Constant_45"
615
+ op_type: "Constant"
616
+ attribute {
617
+ name: "value"
618
+ t {
619
+ data_type: 7
620
+ raw_data: "\001\000\000\000\000\000\000\000"
621
+ }
622
+ type: TENSOR
623
+ }
624
+ }
625
+ node {
626
+ input: "67"
627
+ input: "68"
628
+ output: "69"
629
+ name: "Gather_46"
630
+ op_type: "Gather"
631
+ attribute {
632
+ name: "axis"
633
+ i: 0
634
+ type: INT
635
+ }
636
+ }
637
+ node {
638
+ input: "63"
639
+ output: "70"
640
+ name: "Shape_47"
641
+ op_type: "Shape"
642
+ }
643
+ node {
644
+ output: "71"
645
+ name: "Constant_48"
646
+ op_type: "Constant"
647
+ attribute {
648
+ name: "value"
649
+ t {
650
+ data_type: 7
651
+ raw_data: "\002\000\000\000\000\000\000\000"
652
+ }
653
+ type: TENSOR
654
+ }
655
+ }
656
+ node {
657
+ input: "70"
658
+ input: "71"
659
+ output: "72"
660
+ name: "Gather_49"
661
+ op_type: "Gather"
662
+ attribute {
663
+ name: "axis"
664
+ i: 0
665
+ type: INT
666
+ }
667
+ }
668
+ node {
669
+ output: "73"
670
+ name: "Constant_50"
671
+ op_type: "Constant"
672
+ attribute {
673
+ name: "value"
674
+ t {
675
+ data_type: 7
676
+ raw_data: "\002\000\000\000\000\000\000\000"
677
+ }
678
+ type: TENSOR
679
+ }
680
+ }
681
+ node {
682
+ input: "69"
683
+ input: "73"
684
+ output: "74"
685
+ name: "Div_51"
686
+ op_type: "Div"
687
+ }
688
+ node {
689
+ input: "74"
690
+ output: "75"
691
+ name: "Cast_52"
692
+ op_type: "Cast"
693
+ attribute {
694
+ name: "to"
695
+ i: 7
696
+ type: INT
697
+ }
698
+ }
699
+ node {
700
+ input: "75"
701
+ output: "76"
702
+ name: "Cast_53"
703
+ op_type: "Cast"
704
+ attribute {
705
+ name: "to"
706
+ i: 7
707
+ type: INT
708
+ }
709
+ }
710
+ node {
711
+ output: "77"
712
+ name: "Constant_54"
713
+ op_type: "Constant"
714
+ attribute {
715
+ name: "value"
716
+ t {
717
+ data_type: 7
718
+ raw_data: "\002\000\000\000\000\000\000\000"
719
+ }
720
+ type: TENSOR
721
+ }
722
+ }
723
+ node {
724
+ input: "72"
725
+ input: "77"
726
+ output: "78"
727
+ name: "Mul_55"
728
+ op_type: "Mul"
729
+ }
730
+ node {
731
+ input: "66"
732
+ output: "80"
733
+ name: "Unsqueeze_56"
734
+ op_type: "Unsqueeze"
735
+ attribute {
736
+ name: "axes"
737
+ ints: 0
738
+ type: INTS
739
+ }
740
+ }
741
+ node {
742
+ input: "76"
743
+ output: "82"
744
+ name: "Unsqueeze_57"
745
+ op_type: "Unsqueeze"
746
+ attribute {
747
+ name: "axes"
748
+ ints: 0
749
+ type: INTS
750
+ }
751
+ }
752
+ node {
753
+ input: "72"
754
+ output: "83"
755
+ name: "Unsqueeze_58"
756
+ op_type: "Unsqueeze"
757
+ attribute {
758
+ name: "axes"
759
+ ints: 0
760
+ type: INTS
761
+ }
762
+ }
763
+ node {
764
+ input: "80"
765
+ input: "186"
766
+ input: "82"
767
+ input: "83"
768
+ output: "84"
769
+ name: "Concat_59"
770
+ op_type: "Concat"
771
+ attribute {
772
+ name: "axis"
773
+ i: 0
774
+ type: INT
775
+ }
776
+ }
777
+ node {
778
+ input: "63"
779
+ input: "84"
780
+ output: "85"
781
+ name: "Reshape_60"
782
+ op_type: "Reshape"
783
+ }
784
+ node {
785
+ input: "85"
786
+ output: "86"
787
+ name: "Transpose_61"
788
+ op_type: "Transpose"
789
+ attribute {
790
+ name: "perm"
791
+ ints: 0
792
+ ints: 2
793
+ ints: 3
794
+ ints: 1
795
+ type: INTS
796
+ }
797
+ }
798
+ node {
799
+ input: "66"
800
+ output: "87"
801
+ name: "Unsqueeze_62"
802
+ op_type: "Unsqueeze"
803
+ attribute {
804
+ name: "axes"
805
+ ints: 0
806
+ type: INTS
807
+ }
808
+ }
809
+ node {
810
+ input: "76"
811
+ output: "88"
812
+ name: "Unsqueeze_63"
813
+ op_type: "Unsqueeze"
814
+ attribute {
815
+ name: "axes"
816
+ ints: 0
817
+ type: INTS
818
+ }
819
+ }
820
+ node {
821
+ input: "78"
822
+ output: "89"
823
+ name: "Unsqueeze_64"
824
+ op_type: "Unsqueeze"
825
+ attribute {
826
+ name: "axes"
827
+ ints: 0
828
+ type: INTS
829
+ }
830
+ }
831
+ node {
832
+ input: "87"
833
+ input: "88"
834
+ input: "89"
835
+ output: "90"
836
+ name: "Concat_65"
837
+ op_type: "Concat"
838
+ attribute {
839
+ name: "axis"
840
+ i: 0
841
+ type: INT
842
+ }
843
+ }
844
+ node {
845
+ input: "86"
846
+ input: "90"
847
+ output: "91"
848
+ name: "Reshape_66"
849
+ op_type: "Reshape"
850
+ }
851
+ node {
852
+ input: "91"
853
+ input: "26"
854
+ output: "92"
855
+ name: "Concat_67"
856
+ op_type: "Concat"
857
+ attribute {
858
+ name: "axis"
859
+ i: 1
860
+ type: INT
861
+ }
862
+ }
863
+ node {
864
+ input: "92"
865
+ input: "up_net.1.0.weight"
866
+ input: "up_net.1.0.bias"
867
+ output: "93"
868
+ name: "Conv_68"
869
+ op_type: "Conv"
870
+ attribute {
871
+ name: "dilations"
872
+ ints: 1
873
+ type: INTS
874
+ }
875
+ attribute {
876
+ name: "group"
877
+ i: 1
878
+ type: INT
879
+ }
880
+ attribute {
881
+ name: "kernel_shape"
882
+ ints: 33
883
+ type: INTS
884
+ }
885
+ attribute {
886
+ name: "pads"
887
+ ints: 16
888
+ ints: 16
889
+ type: INTS
890
+ }
891
+ attribute {
892
+ name: "strides"
893
+ ints: 1
894
+ type: INTS
895
+ }
896
+ }
897
+ node {
898
+ input: "93"
899
+ output: "94"
900
+ name: "Relu_69"
901
+ op_type: "Relu"
902
+ }
903
+ node {
904
+ input: "94"
905
+ output: "95"
906
+ name: "Shape_70"
907
+ op_type: "Shape"
908
+ }
909
+ node {
910
+ output: "96"
911
+ name: "Constant_71"
912
+ op_type: "Constant"
913
+ attribute {
914
+ name: "value"
915
+ t {
916
+ data_type: 7
917
+ raw_data: "\000\000\000\000\000\000\000\000"
918
+ }
919
+ type: TENSOR
920
+ }
921
+ }
922
+ node {
923
+ input: "95"
924
+ input: "96"
925
+ output: "97"
926
+ name: "Gather_72"
927
+ op_type: "Gather"
928
+ attribute {
929
+ name: "axis"
930
+ i: 0
931
+ type: INT
932
+ }
933
+ }
934
+ node {
935
+ input: "94"
936
+ output: "98"
937
+ name: "Shape_73"
938
+ op_type: "Shape"
939
+ }
940
+ node {
941
+ output: "99"
942
+ name: "Constant_74"
943
+ op_type: "Constant"
944
+ attribute {
945
+ name: "value"
946
+ t {
947
+ data_type: 7
948
+ raw_data: "\001\000\000\000\000\000\000\000"
949
+ }
950
+ type: TENSOR
951
+ }
952
+ }
953
+ node {
954
+ input: "98"
955
+ input: "99"
956
+ output: "100"
957
+ name: "Gather_75"
958
+ op_type: "Gather"
959
+ attribute {
960
+ name: "axis"
961
+ i: 0
962
+ type: INT
963
+ }
964
+ }
965
+ node {
966
+ input: "94"
967
+ output: "101"
968
+ name: "Shape_76"
969
+ op_type: "Shape"
970
+ }
971
+ node {
972
+ output: "102"
973
+ name: "Constant_77"
974
+ op_type: "Constant"
975
+ attribute {
976
+ name: "value"
977
+ t {
978
+ data_type: 7
979
+ raw_data: "\002\000\000\000\000\000\000\000"
980
+ }
981
+ type: TENSOR
982
+ }
983
+ }
984
+ node {
985
+ input: "101"
986
+ input: "102"
987
+ output: "103"
988
+ name: "Gather_78"
989
+ op_type: "Gather"
990
+ attribute {
991
+ name: "axis"
992
+ i: 0
993
+ type: INT
994
+ }
995
+ }
996
+ node {
997
+ output: "104"
998
+ name: "Constant_79"
999
+ op_type: "Constant"
1000
+ attribute {
1001
+ name: "value"
1002
+ t {
1003
+ data_type: 7
1004
+ raw_data: "\002\000\000\000\000\000\000\000"
1005
+ }
1006
+ type: TENSOR
1007
+ }
1008
+ }
1009
+ node {
1010
+ input: "100"
1011
+ input: "104"
1012
+ output: "105"
1013
+ name: "Div_80"
1014
+ op_type: "Div"
1015
+ }
1016
+ node {
1017
+ input: "105"
1018
+ output: "106"
1019
+ name: "Cast_81"
1020
+ op_type: "Cast"
1021
+ attribute {
1022
+ name: "to"
1023
+ i: 7
1024
+ type: INT
1025
+ }
1026
+ }
1027
+ node {
1028
+ input: "106"
1029
+ output: "107"
1030
+ name: "Cast_82"
1031
+ op_type: "Cast"
1032
+ attribute {
1033
+ name: "to"
1034
+ i: 7
1035
+ type: INT
1036
+ }
1037
+ }
1038
+ node {
1039
+ output: "108"
1040
+ name: "Constant_83"
1041
+ op_type: "Constant"
1042
+ attribute {
1043
+ name: "value"
1044
+ t {
1045
+ data_type: 7
1046
+ raw_data: "\002\000\000\000\000\000\000\000"
1047
+ }
1048
+ type: TENSOR
1049
+ }
1050
+ }
1051
+ node {
1052
+ input: "103"
1053
+ input: "108"
1054
+ output: "109"
1055
+ name: "Mul_84"
1056
+ op_type: "Mul"
1057
+ }
1058
+ node {
1059
+ input: "97"
1060
+ output: "111"
1061
+ name: "Unsqueeze_85"
1062
+ op_type: "Unsqueeze"
1063
+ attribute {
1064
+ name: "axes"
1065
+ ints: 0
1066
+ type: INTS
1067
+ }
1068
+ }
1069
+ node {
1070
+ input: "107"
1071
+ output: "113"
1072
+ name: "Unsqueeze_86"
1073
+ op_type: "Unsqueeze"
1074
+ attribute {
1075
+ name: "axes"
1076
+ ints: 0
1077
+ type: INTS
1078
+ }
1079
+ }
1080
+ node {
1081
+ input: "103"
1082
+ output: "114"
1083
+ name: "Unsqueeze_87"
1084
+ op_type: "Unsqueeze"
1085
+ attribute {
1086
+ name: "axes"
1087
+ ints: 0
1088
+ type: INTS
1089
+ }
1090
+ }
1091
+ node {
1092
+ input: "111"
1093
+ input: "187"
1094
+ input: "113"
1095
+ input: "114"
1096
+ output: "115"
1097
+ name: "Concat_88"
1098
+ op_type: "Concat"
1099
+ attribute {
1100
+ name: "axis"
1101
+ i: 0
1102
+ type: INT
1103
+ }
1104
+ }
1105
+ node {
1106
+ input: "94"
1107
+ input: "115"
1108
+ output: "116"
1109
+ name: "Reshape_89"
1110
+ op_type: "Reshape"
1111
+ }
1112
+ node {
1113
+ input: "116"
1114
+ output: "117"
1115
+ name: "Transpose_90"
1116
+ op_type: "Transpose"
1117
+ attribute {
1118
+ name: "perm"
1119
+ ints: 0
1120
+ ints: 2
1121
+ ints: 3
1122
+ ints: 1
1123
+ type: INTS
1124
+ }
1125
+ }
1126
+ node {
1127
+ input: "97"
1128
+ output: "118"
1129
+ name: "Unsqueeze_91"
1130
+ op_type: "Unsqueeze"
1131
+ attribute {
1132
+ name: "axes"
1133
+ ints: 0
1134
+ type: INTS
1135
+ }
1136
+ }
1137
+ node {
1138
+ input: "107"
1139
+ output: "119"
1140
+ name: "Unsqueeze_92"
1141
+ op_type: "Unsqueeze"
1142
+ attribute {
1143
+ name: "axes"
1144
+ ints: 0
1145
+ type: INTS
1146
+ }
1147
+ }
1148
+ node {
1149
+ input: "109"
1150
+ output: "120"
1151
+ name: "Unsqueeze_93"
1152
+ op_type: "Unsqueeze"
1153
+ attribute {
1154
+ name: "axes"
1155
+ ints: 0
1156
+ type: INTS
1157
+ }
1158
+ }
1159
+ node {
1160
+ input: "118"
1161
+ input: "119"
1162
+ input: "120"
1163
+ output: "121"
1164
+ name: "Concat_94"
1165
+ op_type: "Concat"
1166
+ attribute {
1167
+ name: "axis"
1168
+ i: 0
1169
+ type: INT
1170
+ }
1171
+ }
1172
+ node {
1173
+ input: "117"
1174
+ input: "121"
1175
+ output: "122"
1176
+ name: "Reshape_95"
1177
+ op_type: "Reshape"
1178
+ }
1179
+ node {
1180
+ input: "122"
1181
+ input: "24"
1182
+ output: "123"
1183
+ name: "Concat_96"
1184
+ op_type: "Concat"
1185
+ attribute {
1186
+ name: "axis"
1187
+ i: 1
1188
+ type: INT
1189
+ }
1190
+ }
1191
+ node {
1192
+ input: "123"
1193
+ input: "up_net.2.0.weight"
1194
+ input: "up_net.2.0.bias"
1195
+ output: "124"
1196
+ name: "Conv_97"
1197
+ op_type: "Conv"
1198
+ attribute {
1199
+ name: "dilations"
1200
+ ints: 1
1201
+ type: INTS
1202
+ }
1203
+ attribute {
1204
+ name: "group"
1205
+ i: 1
1206
+ type: INT
1207
+ }
1208
+ attribute {
1209
+ name: "kernel_shape"
1210
+ ints: 65
1211
+ type: INTS
1212
+ }
1213
+ attribute {
1214
+ name: "pads"
1215
+ ints: 32
1216
+ ints: 32
1217
+ type: INTS
1218
+ }
1219
+ attribute {
1220
+ name: "strides"
1221
+ ints: 1
1222
+ type: INTS
1223
+ }
1224
+ }
1225
+ node {
1226
+ input: "124"
1227
+ output: "125"
1228
+ name: "Relu_98"
1229
+ op_type: "Relu"
1230
+ }
1231
+ node {
1232
+ input: "125"
1233
+ output: "126"
1234
+ name: "Shape_99"
1235
+ op_type: "Shape"
1236
+ }
1237
+ node {
1238
+ output: "127"
1239
+ name: "Constant_100"
1240
+ op_type: "Constant"
1241
+ attribute {
1242
+ name: "value"
1243
+ t {
1244
+ data_type: 7
1245
+ raw_data: "\000\000\000\000\000\000\000\000"
1246
+ }
1247
+ type: TENSOR
1248
+ }
1249
+ }
1250
+ node {
1251
+ input: "126"
1252
+ input: "127"
1253
+ output: "128"
1254
+ name: "Gather_101"
1255
+ op_type: "Gather"
1256
+ attribute {
1257
+ name: "axis"
1258
+ i: 0
1259
+ type: INT
1260
+ }
1261
+ }
1262
+ node {
1263
+ input: "125"
1264
+ output: "129"
1265
+ name: "Shape_102"
1266
+ op_type: "Shape"
1267
+ }
1268
+ node {
1269
+ output: "130"
1270
+ name: "Constant_103"
1271
+ op_type: "Constant"
1272
+ attribute {
1273
+ name: "value"
1274
+ t {
1275
+ data_type: 7
1276
+ raw_data: "\001\000\000\000\000\000\000\000"
1277
+ }
1278
+ type: TENSOR
1279
+ }
1280
+ }
1281
+ node {
1282
+ input: "129"
1283
+ input: "130"
1284
+ output: "131"
1285
+ name: "Gather_104"
1286
+ op_type: "Gather"
1287
+ attribute {
1288
+ name: "axis"
1289
+ i: 0
1290
+ type: INT
1291
+ }
1292
+ }
1293
+ node {
1294
+ input: "125"
1295
+ output: "132"
1296
+ name: "Shape_105"
1297
+ op_type: "Shape"
1298
+ }
1299
+ node {
1300
+ output: "133"
1301
+ name: "Constant_106"
1302
+ op_type: "Constant"
1303
+ attribute {
1304
+ name: "value"
1305
+ t {
1306
+ data_type: 7
1307
+ raw_data: "\002\000\000\000\000\000\000\000"
1308
+ }
1309
+ type: TENSOR
1310
+ }
1311
+ }
1312
+ node {
1313
+ input: "132"
1314
+ input: "133"
1315
+ output: "134"
1316
+ name: "Gather_107"
1317
+ op_type: "Gather"
1318
+ attribute {
1319
+ name: "axis"
1320
+ i: 0
1321
+ type: INT
1322
+ }
1323
+ }
1324
+ node {
1325
+ output: "135"
1326
+ name: "Constant_108"
1327
+ op_type: "Constant"
1328
+ attribute {
1329
+ name: "value"
1330
+ t {
1331
+ data_type: 7
1332
+ raw_data: "\002\000\000\000\000\000\000\000"
1333
+ }
1334
+ type: TENSOR
1335
+ }
1336
+ }
1337
+ node {
1338
+ input: "131"
1339
+ input: "135"
1340
+ output: "136"
1341
+ name: "Div_109"
1342
+ op_type: "Div"
1343
+ }
1344
+ node {
1345
+ input: "136"
1346
+ output: "137"
1347
+ name: "Cast_110"
1348
+ op_type: "Cast"
1349
+ attribute {
1350
+ name: "to"
1351
+ i: 7
1352
+ type: INT
1353
+ }
1354
+ }
1355
+ node {
1356
+ input: "137"
1357
+ output: "138"
1358
+ name: "Cast_111"
1359
+ op_type: "Cast"
1360
+ attribute {
1361
+ name: "to"
1362
+ i: 7
1363
+ type: INT
1364
+ }
1365
+ }
1366
+ node {
1367
+ output: "139"
1368
+ name: "Constant_112"
1369
+ op_type: "Constant"
1370
+ attribute {
1371
+ name: "value"
1372
+ t {
1373
+ data_type: 7
1374
+ raw_data: "\002\000\000\000\000\000\000\000"
1375
+ }
1376
+ type: TENSOR
1377
+ }
1378
+ }
1379
+ node {
1380
+ input: "134"
1381
+ input: "139"
1382
+ output: "140"
1383
+ name: "Mul_113"
1384
+ op_type: "Mul"
1385
+ }
1386
+ node {
1387
+ input: "128"
1388
+ output: "142"
1389
+ name: "Unsqueeze_114"
1390
+ op_type: "Unsqueeze"
1391
+ attribute {
1392
+ name: "axes"
1393
+ ints: 0
1394
+ type: INTS
1395
+ }
1396
+ }
1397
+ node {
1398
+ input: "138"
1399
+ output: "144"
1400
+ name: "Unsqueeze_115"
1401
+ op_type: "Unsqueeze"
1402
+ attribute {
1403
+ name: "axes"
1404
+ ints: 0
1405
+ type: INTS
1406
+ }
1407
+ }
1408
+ node {
1409
+ input: "134"
1410
+ output: "145"
1411
+ name: "Unsqueeze_116"
1412
+ op_type: "Unsqueeze"
1413
+ attribute {
1414
+ name: "axes"
1415
+ ints: 0
1416
+ type: INTS
1417
+ }
1418
+ }
1419
+ node {
1420
+ input: "142"
1421
+ input: "188"
1422
+ input: "144"
1423
+ input: "145"
1424
+ output: "146"
1425
+ name: "Concat_117"
1426
+ op_type: "Concat"
1427
+ attribute {
1428
+ name: "axis"
1429
+ i: 0
1430
+ type: INT
1431
+ }
1432
+ }
1433
+ node {
1434
+ input: "125"
1435
+ input: "146"
1436
+ output: "147"
1437
+ name: "Reshape_118"
1438
+ op_type: "Reshape"
1439
+ }
1440
+ node {
1441
+ input: "147"
1442
+ output: "148"
1443
+ name: "Transpose_119"
1444
+ op_type: "Transpose"
1445
+ attribute {
1446
+ name: "perm"
1447
+ ints: 0
1448
+ ints: 2
1449
+ ints: 3
1450
+ ints: 1
1451
+ type: INTS
1452
+ }
1453
+ }
1454
+ node {
1455
+ input: "128"
1456
+ output: "149"
1457
+ name: "Unsqueeze_120"
1458
+ op_type: "Unsqueeze"
1459
+ attribute {
1460
+ name: "axes"
1461
+ ints: 0
1462
+ type: INTS
1463
+ }
1464
+ }
1465
+ node {
1466
+ input: "138"
1467
+ output: "150"
1468
+ name: "Unsqueeze_121"
1469
+ op_type: "Unsqueeze"
1470
+ attribute {
1471
+ name: "axes"
1472
+ ints: 0
1473
+ type: INTS
1474
+ }
1475
+ }
1476
+ node {
1477
+ input: "140"
1478
+ output: "151"
1479
+ name: "Unsqueeze_122"
1480
+ op_type: "Unsqueeze"
1481
+ attribute {
1482
+ name: "axes"
1483
+ ints: 0
1484
+ type: INTS
1485
+ }
1486
+ }
1487
+ node {
1488
+ input: "149"
1489
+ input: "150"
1490
+ input: "151"
1491
+ output: "152"
1492
+ name: "Concat_123"
1493
+ op_type: "Concat"
1494
+ attribute {
1495
+ name: "axis"
1496
+ i: 0
1497
+ type: INT
1498
+ }
1499
+ }
1500
+ node {
1501
+ input: "148"
1502
+ input: "152"
1503
+ output: "153"
1504
+ name: "Reshape_124"
1505
+ op_type: "Reshape"
1506
+ }
1507
+ node {
1508
+ input: "153"
1509
+ input: "22"
1510
+ output: "154"
1511
+ name: "Concat_125"
1512
+ op_type: "Concat"
1513
+ attribute {
1514
+ name: "axis"
1515
+ i: 1
1516
+ type: INT
1517
+ }
1518
+ }
1519
+ node {
1520
+ input: "154"
1521
+ input: "up_net.3.0.weight"
1522
+ input: "up_net.3.0.bias"
1523
+ output: "155"
1524
+ name: "Conv_126"
1525
+ op_type: "Conv"
1526
+ attribute {
1527
+ name: "dilations"
1528
+ ints: 1
1529
+ type: INTS
1530
+ }
1531
+ attribute {
1532
+ name: "group"
1533
+ i: 1
1534
+ type: INT
1535
+ }
1536
+ attribute {
1537
+ name: "kernel_shape"
1538
+ ints: 9
1539
+ type: INTS
1540
+ }
1541
+ attribute {
1542
+ name: "pads"
1543
+ ints: 4
1544
+ ints: 4
1545
+ type: INTS
1546
+ }
1547
+ attribute {
1548
+ name: "strides"
1549
+ ints: 1
1550
+ type: INTS
1551
+ }
1552
+ }
1553
+ node {
1554
+ input: "155"
1555
+ output: "156"
1556
+ name: "Shape_127"
1557
+ op_type: "Shape"
1558
+ }
1559
+ node {
1560
+ output: "157"
1561
+ name: "Constant_128"
1562
+ op_type: "Constant"
1563
+ attribute {
1564
+ name: "value"
1565
+ t {
1566
+ data_type: 7
1567
+ raw_data: "\000\000\000\000\000\000\000\000"
1568
+ }
1569
+ type: TENSOR
1570
+ }
1571
+ }
1572
+ node {
1573
+ input: "156"
1574
+ input: "157"
1575
+ output: "158"
1576
+ name: "Gather_129"
1577
+ op_type: "Gather"
1578
+ attribute {
1579
+ name: "axis"
1580
+ i: 0
1581
+ type: INT
1582
+ }
1583
+ }
1584
+ node {
1585
+ input: "155"
1586
+ output: "159"
1587
+ name: "Shape_130"
1588
+ op_type: "Shape"
1589
+ }
1590
+ node {
1591
+ output: "160"
1592
+ name: "Constant_131"
1593
+ op_type: "Constant"
1594
+ attribute {
1595
+ name: "value"
1596
+ t {
1597
+ data_type: 7
1598
+ raw_data: "\001\000\000\000\000\000\000\000"
1599
+ }
1600
+ type: TENSOR
1601
+ }
1602
+ }
1603
+ node {
1604
+ input: "159"
1605
+ input: "160"
1606
+ output: "161"
1607
+ name: "Gather_132"
1608
+ op_type: "Gather"
1609
+ attribute {
1610
+ name: "axis"
1611
+ i: 0
1612
+ type: INT
1613
+ }
1614
+ }
1615
+ node {
1616
+ input: "155"
1617
+ output: "162"
1618
+ name: "Shape_133"
1619
+ op_type: "Shape"
1620
+ }
1621
+ node {
1622
+ output: "163"
1623
+ name: "Constant_134"
1624
+ op_type: "Constant"
1625
+ attribute {
1626
+ name: "value"
1627
+ t {
1628
+ data_type: 7
1629
+ raw_data: "\002\000\000\000\000\000\000\000"
1630
+ }
1631
+ type: TENSOR
1632
+ }
1633
+ }
1634
+ node {
1635
+ input: "162"
1636
+ input: "163"
1637
+ output: "164"
1638
+ name: "Gather_135"
1639
+ op_type: "Gather"
1640
+ attribute {
1641
+ name: "axis"
1642
+ i: 0
1643
+ type: INT
1644
+ }
1645
+ }
1646
+ node {
1647
+ output: "165"
1648
+ name: "Constant_136"
1649
+ op_type: "Constant"
1650
+ attribute {
1651
+ name: "value"
1652
+ t {
1653
+ data_type: 7
1654
+ raw_data: "\002\000\000\000\000\000\000\000"
1655
+ }
1656
+ type: TENSOR
1657
+ }
1658
+ }
1659
+ node {
1660
+ input: "161"
1661
+ input: "165"
1662
+ output: "166"
1663
+ name: "Div_137"
1664
+ op_type: "Div"
1665
+ }
1666
+ node {
1667
+ input: "166"
1668
+ output: "167"
1669
+ name: "Cast_138"
1670
+ op_type: "Cast"
1671
+ attribute {
1672
+ name: "to"
1673
+ i: 7
1674
+ type: INT
1675
+ }
1676
+ }
1677
+ node {
1678
+ input: "167"
1679
+ output: "168"
1680
+ name: "Cast_139"
1681
+ op_type: "Cast"
1682
+ attribute {
1683
+ name: "to"
1684
+ i: 7
1685
+ type: INT
1686
+ }
1687
+ }
1688
+ node {
1689
+ output: "169"
1690
+ name: "Constant_140"
1691
+ op_type: "Constant"
1692
+ attribute {
1693
+ name: "value"
1694
+ t {
1695
+ data_type: 7
1696
+ raw_data: "\002\000\000\000\000\000\000\000"
1697
+ }
1698
+ type: TENSOR
1699
+ }
1700
+ }
1701
+ node {
1702
+ input: "164"
1703
+ input: "169"
1704
+ output: "170"
1705
+ name: "Mul_141"
1706
+ op_type: "Mul"
1707
+ }
1708
+ node {
1709
+ input: "158"
1710
+ output: "172"
1711
+ name: "Unsqueeze_142"
1712
+ op_type: "Unsqueeze"
1713
+ attribute {
1714
+ name: "axes"
1715
+ ints: 0
1716
+ type: INTS
1717
+ }
1718
+ }
1719
+ node {
1720
+ input: "168"
1721
+ output: "174"
1722
+ name: "Unsqueeze_143"
1723
+ op_type: "Unsqueeze"
1724
+ attribute {
1725
+ name: "axes"
1726
+ ints: 0
1727
+ type: INTS
1728
+ }
1729
+ }
1730
+ node {
1731
+ input: "164"
1732
+ output: "175"
1733
+ name: "Unsqueeze_144"
1734
+ op_type: "Unsqueeze"
1735
+ attribute {
1736
+ name: "axes"
1737
+ ints: 0
1738
+ type: INTS
1739
+ }
1740
+ }
1741
+ node {
1742
+ input: "172"
1743
+ input: "189"
1744
+ input: "174"
1745
+ input: "175"
1746
+ output: "176"
1747
+ name: "Concat_145"
1748
+ op_type: "Concat"
1749
+ attribute {
1750
+ name: "axis"
1751
+ i: 0
1752
+ type: INT
1753
+ }
1754
+ }
1755
+ node {
1756
+ input: "155"
1757
+ input: "176"
1758
+ output: "177"
1759
+ name: "Reshape_146"
1760
+ op_type: "Reshape"
1761
+ }
1762
+ node {
1763
+ input: "177"
1764
+ output: "178"
1765
+ name: "Transpose_147"
1766
+ op_type: "Transpose"
1767
+ attribute {
1768
+ name: "perm"
1769
+ ints: 0
1770
+ ints: 2
1771
+ ints: 3
1772
+ ints: 1
1773
+ type: INTS
1774
+ }
1775
+ }
1776
+ node {
1777
+ input: "158"
1778
+ output: "179"
1779
+ name: "Unsqueeze_148"
1780
+ op_type: "Unsqueeze"
1781
+ attribute {
1782
+ name: "axes"
1783
+ ints: 0
1784
+ type: INTS
1785
+ }
1786
+ }
1787
+ node {
1788
+ input: "168"
1789
+ output: "180"
1790
+ name: "Unsqueeze_149"
1791
+ op_type: "Unsqueeze"
1792
+ attribute {
1793
+ name: "axes"
1794
+ ints: 0
1795
+ type: INTS
1796
+ }
1797
+ }
1798
+ node {
1799
+ input: "170"
1800
+ output: "181"
1801
+ name: "Unsqueeze_150"
1802
+ op_type: "Unsqueeze"
1803
+ attribute {
1804
+ name: "axes"
1805
+ ints: 0
1806
+ type: INTS
1807
+ }
1808
+ }
1809
+ node {
1810
+ input: "179"
1811
+ input: "180"
1812
+ input: "181"
1813
+ output: "182"
1814
+ name: "Concat_151"
1815
+ op_type: "Concat"
1816
+ attribute {
1817
+ name: "axis"
1818
+ i: 0
1819
+ type: INT
1820
+ }
1821
+ }
1822
+ node {
1823
+ input: "178"
1824
+ input: "182"
1825
+ output: "183"
1826
+ name: "Reshape_152"
1827
+ op_type: "Reshape"
1828
+ }
1829
+ node {
1830
+ input: "183"
1831
+ input: "x"
1832
+ output: "y"
1833
+ name: "Add_153"
1834
+ op_type: "Add"
1835
+ }
1836
+ initializer {
1837
+ dims: 1
1838
+ data_type: 7
1839
+ name: "185"
1840
+ raw_data: "\002\000\000\000\000\000\000\000"
1841
+ }
1842
+ initializer {
1843
+ dims: 1
1844
+ data_type: 7
1845
+ name: "186"
1846
+ raw_data: "\002\000\000\000\000\000\000\000"
1847
+ }
1848
+ initializer {
1849
+ dims: 1
1850
+ data_type: 7
1851
+ name: "187"
1852
+ raw_data: "\002\000\000\000\000\000\000\000"
1853
+ }
1854
+ initializer {
1855
+ dims: 1
1856
+ data_type: 7
1857
+ name: "188"
1858
+ raw_data: "\002\000\000\000\000\000\000\000"
1859
+ }
1860
+ initializer {
1861
+ dims: 1
1862
+ data_type: 7
1863
+ name: "189"
1864
+ raw_data: "\002\000\000\000\000\000\000\000"
1865
+ }
1866
+ initializer {
1867
+ dims: 512
1868
+ data_type: 1
1869
+ name: "bottleneck.0.bias"
1870
+ }
1871
+ initializer {
1872
+ dims: 512
1873
+ dims: 512
1874
+ dims: 9
1875
+ data_type: 1
1876
+ name: "bottleneck.0.weight"
1877
+ }
1878
+ initializer {
1879
+ dims: 1024
1880
+ data_type: 1
1881
+ name: "bottleneck.2.bias"
1882
+ }
1883
+ initializer {
1884
+ dims: 1024
1885
+ dims: 512
1886
+ dims: 9
1887
+ data_type: 1
1888
+ name: "bottleneck.2.weight"
1889
+ }
1890
+ initializer {
1891
+ dims: 128
1892
+ data_type: 1
1893
+ name: "down_net.0.0.bias"
1894
+ }
1895
+ initializer {
1896
+ dims: 128
1897
+ dims: 2
1898
+ dims: 65
1899
+ data_type: 1
1900
+ name: "down_net.0.0.weight"
1901
+ }
1902
+ initializer {
1903
+ dims: 256
1904
+ data_type: 1
1905
+ name: "down_net.1.0.bias"
1906
+ }
1907
+ initializer {
1908
+ dims: 256
1909
+ dims: 128
1910
+ dims: 33
1911
+ data_type: 1
1912
+ name: "down_net.1.0.weight"
1913
+ }
1914
+ initializer {
1915
+ dims: 512
1916
+ data_type: 1
1917
+ name: "down_net.2.0.bias"
1918
+ }
1919
+ initializer {
1920
+ dims: 512
1921
+ dims: 256
1922
+ dims: 17
1923
+ data_type: 1
1924
+ name: "down_net.2.0.weight"
1925
+ }
1926
+ initializer {
1927
+ dims: 512
1928
+ data_type: 1
1929
+ name: "down_net.3.0.bias"
1930
+ }
1931
+ initializer {
1932
+ dims: 512
1933
+ dims: 512
1934
+ dims: 9
1935
+ data_type: 1
1936
+ name: "down_net.3.0.weight"
1937
+ }
1938
+ initializer {
1939
+ dims: 1024
1940
+ data_type: 1
1941
+ name: "up_net.0.0.bias"
1942
+ }
1943
+ initializer {
1944
+ dims: 1024
1945
+ dims: 1024
1946
+ dims: 17
1947
+ data_type: 1
1948
+ name: "up_net.0.0.weight"
1949
+ }
1950
+ initializer {
1951
+ dims: 512
1952
+ data_type: 1
1953
+ name: "up_net.1.0.bias"
1954
+ }
1955
+ initializer {
1956
+ dims: 512
1957
+ dims: 1024
1958
+ dims: 33
1959
+ data_type: 1
1960
+ name: "up_net.1.0.weight"
1961
+ }
1962
+ initializer {
1963
+ dims: 256
1964
+ data_type: 1
1965
+ name: "up_net.2.0.bias"
1966
+ }
1967
+ initializer {
1968
+ dims: 256
1969
+ dims: 512
1970
+ dims: 65
1971
+ data_type: 1
1972
+ name: "up_net.2.0.weight"
1973
+ }
1974
+ initializer {
1975
+ dims: 4
1976
+ data_type: 1
1977
+ name: "up_net.3.0.bias"
1978
+ raw_data: "\276V%;o\026e<\301\334#<\250z\034\273"
1979
+ }
1980
+ initializer {
1981
+ dims: 4
1982
+ dims: 256
1983
+ dims: 9
1984
+ data_type: 1
1985
+ name: "up_net.3.0.weight"
1986
+ }
1987
+ input {
1988
+ name: "x"
1989
+ type {
1990
+ tensor_type {
1991
+ elem_type: 1
1992
+ shape {
1993
+ dim {
1994
+ dim_param: "batch_size"
1995
+ }
1996
+ dim {
1997
+ dim_value: 2
1998
+ }
1999
+ dim {
2000
+ dim_value: 8192
2001
+ }
2002
+ }
2003
+ }
2004
+ }
2005
+ }
2006
+ output {
2007
+ name: "y"
2008
+ type {
2009
+ tensor_type {
2010
+ elem_type: 1
2011
+ shape {
2012
+ dim {
2013
+ dim_param: "batch_size"
2014
+ }
2015
+ dim {
2016
+ dim_value: 2
2017
+ }
2018
+ dim {
2019
+ dim_value: 8192
2020
+ }
2021
+ }
2022
+ }
2023
+ }
2024
+ }
2025
+ }
2026
+ opset_import {
2027
+ version: 11
2028
+ }
models/ailia-models/unetdo.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f16be7ee903b1a4e33e5a34e04f6337958088c3d817fef769aba3c6435cb3b2a
3
+ size 225706017
models/ailia-models/unetdo.onnx.prototxt ADDED
@@ -0,0 +1,2028 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ir_version: 6
2
+ producer_name: "pytorch"
3
+ producer_version: "1.7"
4
+ model_version: 0
5
+ graph {
6
+ name: "torch-jit-export"
7
+ node {
8
+ input: "x"
9
+ input: "down_net.0.0.weight"
10
+ input: "down_net.0.0.bias"
11
+ output: "21"
12
+ name: "Conv_0"
13
+ op_type: "Conv"
14
+ attribute {
15
+ name: "dilations"
16
+ ints: 1
17
+ type: INTS
18
+ }
19
+ attribute {
20
+ name: "group"
21
+ i: 1
22
+ type: INT
23
+ }
24
+ attribute {
25
+ name: "kernel_shape"
26
+ ints: 65
27
+ type: INTS
28
+ }
29
+ attribute {
30
+ name: "pads"
31
+ ints: 32
32
+ ints: 32
33
+ type: INTS
34
+ }
35
+ attribute {
36
+ name: "strides"
37
+ ints: 2
38
+ type: INTS
39
+ }
40
+ }
41
+ node {
42
+ input: "21"
43
+ output: "22"
44
+ name: "Relu_1"
45
+ op_type: "Relu"
46
+ }
47
+ node {
48
+ input: "22"
49
+ input: "down_net.1.0.weight"
50
+ input: "down_net.1.0.bias"
51
+ output: "23"
52
+ name: "Conv_2"
53
+ op_type: "Conv"
54
+ attribute {
55
+ name: "dilations"
56
+ ints: 1
57
+ type: INTS
58
+ }
59
+ attribute {
60
+ name: "group"
61
+ i: 1
62
+ type: INT
63
+ }
64
+ attribute {
65
+ name: "kernel_shape"
66
+ ints: 33
67
+ type: INTS
68
+ }
69
+ attribute {
70
+ name: "pads"
71
+ ints: 16
72
+ ints: 16
73
+ type: INTS
74
+ }
75
+ attribute {
76
+ name: "strides"
77
+ ints: 2
78
+ type: INTS
79
+ }
80
+ }
81
+ node {
82
+ input: "23"
83
+ output: "24"
84
+ name: "Relu_3"
85
+ op_type: "Relu"
86
+ }
87
+ node {
88
+ input: "24"
89
+ input: "down_net.2.0.weight"
90
+ input: "down_net.2.0.bias"
91
+ output: "25"
92
+ name: "Conv_4"
93
+ op_type: "Conv"
94
+ attribute {
95
+ name: "dilations"
96
+ ints: 1
97
+ type: INTS
98
+ }
99
+ attribute {
100
+ name: "group"
101
+ i: 1
102
+ type: INT
103
+ }
104
+ attribute {
105
+ name: "kernel_shape"
106
+ ints: 17
107
+ type: INTS
108
+ }
109
+ attribute {
110
+ name: "pads"
111
+ ints: 8
112
+ ints: 8
113
+ type: INTS
114
+ }
115
+ attribute {
116
+ name: "strides"
117
+ ints: 2
118
+ type: INTS
119
+ }
120
+ }
121
+ node {
122
+ input: "25"
123
+ output: "26"
124
+ name: "Relu_5"
125
+ op_type: "Relu"
126
+ }
127
+ node {
128
+ input: "26"
129
+ input: "down_net.3.0.weight"
130
+ input: "down_net.3.0.bias"
131
+ output: "27"
132
+ name: "Conv_6"
133
+ op_type: "Conv"
134
+ attribute {
135
+ name: "dilations"
136
+ ints: 1
137
+ type: INTS
138
+ }
139
+ attribute {
140
+ name: "group"
141
+ i: 1
142
+ type: INT
143
+ }
144
+ attribute {
145
+ name: "kernel_shape"
146
+ ints: 9
147
+ type: INTS
148
+ }
149
+ attribute {
150
+ name: "pads"
151
+ ints: 4
152
+ ints: 4
153
+ type: INTS
154
+ }
155
+ attribute {
156
+ name: "strides"
157
+ ints: 2
158
+ type: INTS
159
+ }
160
+ }
161
+ node {
162
+ input: "27"
163
+ output: "28"
164
+ name: "Relu_7"
165
+ op_type: "Relu"
166
+ }
167
+ node {
168
+ input: "28"
169
+ input: "bottleneck.0.weight"
170
+ input: "bottleneck.0.bias"
171
+ output: "29"
172
+ name: "Conv_8"
173
+ op_type: "Conv"
174
+ attribute {
175
+ name: "dilations"
176
+ ints: 1
177
+ type: INTS
178
+ }
179
+ attribute {
180
+ name: "group"
181
+ i: 1
182
+ type: INT
183
+ }
184
+ attribute {
185
+ name: "kernel_shape"
186
+ ints: 9
187
+ type: INTS
188
+ }
189
+ attribute {
190
+ name: "pads"
191
+ ints: 4
192
+ ints: 4
193
+ type: INTS
194
+ }
195
+ attribute {
196
+ name: "strides"
197
+ ints: 2
198
+ type: INTS
199
+ }
200
+ }
201
+ node {
202
+ input: "29"
203
+ output: "30"
204
+ name: "Relu_9"
205
+ op_type: "Relu"
206
+ }
207
+ node {
208
+ input: "30"
209
+ input: "bottleneck.3.weight"
210
+ input: "bottleneck.3.bias"
211
+ output: "31"
212
+ name: "Conv_10"
213
+ op_type: "Conv"
214
+ attribute {
215
+ name: "dilations"
216
+ ints: 1
217
+ type: INTS
218
+ }
219
+ attribute {
220
+ name: "group"
221
+ i: 1
222
+ type: INT
223
+ }
224
+ attribute {
225
+ name: "kernel_shape"
226
+ ints: 9
227
+ type: INTS
228
+ }
229
+ attribute {
230
+ name: "pads"
231
+ ints: 4
232
+ ints: 4
233
+ type: INTS
234
+ }
235
+ attribute {
236
+ name: "strides"
237
+ ints: 1
238
+ type: INTS
239
+ }
240
+ }
241
+ node {
242
+ input: "31"
243
+ output: "32"
244
+ name: "Relu_11"
245
+ op_type: "Relu"
246
+ }
247
+ node {
248
+ input: "32"
249
+ output: "33"
250
+ name: "Shape_12"
251
+ op_type: "Shape"
252
+ }
253
+ node {
254
+ output: "34"
255
+ name: "Constant_13"
256
+ op_type: "Constant"
257
+ attribute {
258
+ name: "value"
259
+ t {
260
+ data_type: 7
261
+ raw_data: "\000\000\000\000\000\000\000\000"
262
+ }
263
+ type: TENSOR
264
+ }
265
+ }
266
+ node {
267
+ input: "33"
268
+ input: "34"
269
+ output: "35"
270
+ name: "Gather_14"
271
+ op_type: "Gather"
272
+ attribute {
273
+ name: "axis"
274
+ i: 0
275
+ type: INT
276
+ }
277
+ }
278
+ node {
279
+ input: "32"
280
+ output: "36"
281
+ name: "Shape_15"
282
+ op_type: "Shape"
283
+ }
284
+ node {
285
+ output: "37"
286
+ name: "Constant_16"
287
+ op_type: "Constant"
288
+ attribute {
289
+ name: "value"
290
+ t {
291
+ data_type: 7
292
+ raw_data: "\001\000\000\000\000\000\000\000"
293
+ }
294
+ type: TENSOR
295
+ }
296
+ }
297
+ node {
298
+ input: "36"
299
+ input: "37"
300
+ output: "38"
301
+ name: "Gather_17"
302
+ op_type: "Gather"
303
+ attribute {
304
+ name: "axis"
305
+ i: 0
306
+ type: INT
307
+ }
308
+ }
309
+ node {
310
+ input: "32"
311
+ output: "39"
312
+ name: "Shape_18"
313
+ op_type: "Shape"
314
+ }
315
+ node {
316
+ output: "40"
317
+ name: "Constant_19"
318
+ op_type: "Constant"
319
+ attribute {
320
+ name: "value"
321
+ t {
322
+ data_type: 7
323
+ raw_data: "\002\000\000\000\000\000\000\000"
324
+ }
325
+ type: TENSOR
326
+ }
327
+ }
328
+ node {
329
+ input: "39"
330
+ input: "40"
331
+ output: "41"
332
+ name: "Gather_20"
333
+ op_type: "Gather"
334
+ attribute {
335
+ name: "axis"
336
+ i: 0
337
+ type: INT
338
+ }
339
+ }
340
+ node {
341
+ output: "42"
342
+ name: "Constant_21"
343
+ op_type: "Constant"
344
+ attribute {
345
+ name: "value"
346
+ t {
347
+ data_type: 7
348
+ raw_data: "\002\000\000\000\000\000\000\000"
349
+ }
350
+ type: TENSOR
351
+ }
352
+ }
353
+ node {
354
+ input: "38"
355
+ input: "42"
356
+ output: "43"
357
+ name: "Div_22"
358
+ op_type: "Div"
359
+ }
360
+ node {
361
+ input: "43"
362
+ output: "44"
363
+ name: "Cast_23"
364
+ op_type: "Cast"
365
+ attribute {
366
+ name: "to"
367
+ i: 7
368
+ type: INT
369
+ }
370
+ }
371
+ node {
372
+ input: "44"
373
+ output: "45"
374
+ name: "Cast_24"
375
+ op_type: "Cast"
376
+ attribute {
377
+ name: "to"
378
+ i: 7
379
+ type: INT
380
+ }
381
+ }
382
+ node {
383
+ output: "46"
384
+ name: "Constant_25"
385
+ op_type: "Constant"
386
+ attribute {
387
+ name: "value"
388
+ t {
389
+ data_type: 7
390
+ raw_data: "\002\000\000\000\000\000\000\000"
391
+ }
392
+ type: TENSOR
393
+ }
394
+ }
395
+ node {
396
+ input: "41"
397
+ input: "46"
398
+ output: "47"
399
+ name: "Mul_26"
400
+ op_type: "Mul"
401
+ }
402
+ node {
403
+ input: "35"
404
+ output: "49"
405
+ name: "Unsqueeze_27"
406
+ op_type: "Unsqueeze"
407
+ attribute {
408
+ name: "axes"
409
+ ints: 0
410
+ type: INTS
411
+ }
412
+ }
413
+ node {
414
+ input: "45"
415
+ output: "51"
416
+ name: "Unsqueeze_28"
417
+ op_type: "Unsqueeze"
418
+ attribute {
419
+ name: "axes"
420
+ ints: 0
421
+ type: INTS
422
+ }
423
+ }
424
+ node {
425
+ input: "41"
426
+ output: "52"
427
+ name: "Unsqueeze_29"
428
+ op_type: "Unsqueeze"
429
+ attribute {
430
+ name: "axes"
431
+ ints: 0
432
+ type: INTS
433
+ }
434
+ }
435
+ node {
436
+ input: "49"
437
+ input: "185"
438
+ input: "51"
439
+ input: "52"
440
+ output: "53"
441
+ name: "Concat_30"
442
+ op_type: "Concat"
443
+ attribute {
444
+ name: "axis"
445
+ i: 0
446
+ type: INT
447
+ }
448
+ }
449
+ node {
450
+ input: "32"
451
+ input: "53"
452
+ output: "54"
453
+ name: "Reshape_31"
454
+ op_type: "Reshape"
455
+ }
456
+ node {
457
+ input: "54"
458
+ output: "55"
459
+ name: "Transpose_32"
460
+ op_type: "Transpose"
461
+ attribute {
462
+ name: "perm"
463
+ ints: 0
464
+ ints: 2
465
+ ints: 3
466
+ ints: 1
467
+ type: INTS
468
+ }
469
+ }
470
+ node {
471
+ input: "35"
472
+ output: "56"
473
+ name: "Unsqueeze_33"
474
+ op_type: "Unsqueeze"
475
+ attribute {
476
+ name: "axes"
477
+ ints: 0
478
+ type: INTS
479
+ }
480
+ }
481
+ node {
482
+ input: "45"
483
+ output: "57"
484
+ name: "Unsqueeze_34"
485
+ op_type: "Unsqueeze"
486
+ attribute {
487
+ name: "axes"
488
+ ints: 0
489
+ type: INTS
490
+ }
491
+ }
492
+ node {
493
+ input: "47"
494
+ output: "58"
495
+ name: "Unsqueeze_35"
496
+ op_type: "Unsqueeze"
497
+ attribute {
498
+ name: "axes"
499
+ ints: 0
500
+ type: INTS
501
+ }
502
+ }
503
+ node {
504
+ input: "56"
505
+ input: "57"
506
+ input: "58"
507
+ output: "59"
508
+ name: "Concat_36"
509
+ op_type: "Concat"
510
+ attribute {
511
+ name: "axis"
512
+ i: 0
513
+ type: INT
514
+ }
515
+ }
516
+ node {
517
+ input: "55"
518
+ input: "59"
519
+ output: "60"
520
+ name: "Reshape_37"
521
+ op_type: "Reshape"
522
+ }
523
+ node {
524
+ input: "60"
525
+ input: "28"
526
+ output: "61"
527
+ name: "Concat_38"
528
+ op_type: "Concat"
529
+ attribute {
530
+ name: "axis"
531
+ i: 1
532
+ type: INT
533
+ }
534
+ }
535
+ node {
536
+ input: "61"
537
+ input: "up_net.0.0.weight"
538
+ input: "up_net.0.0.bias"
539
+ output: "62"
540
+ name: "Conv_39"
541
+ op_type: "Conv"
542
+ attribute {
543
+ name: "dilations"
544
+ ints: 1
545
+ type: INTS
546
+ }
547
+ attribute {
548
+ name: "group"
549
+ i: 1
550
+ type: INT
551
+ }
552
+ attribute {
553
+ name: "kernel_shape"
554
+ ints: 17
555
+ type: INTS
556
+ }
557
+ attribute {
558
+ name: "pads"
559
+ ints: 8
560
+ ints: 8
561
+ type: INTS
562
+ }
563
+ attribute {
564
+ name: "strides"
565
+ ints: 1
566
+ type: INTS
567
+ }
568
+ }
569
+ node {
570
+ input: "62"
571
+ output: "63"
572
+ name: "Relu_40"
573
+ op_type: "Relu"
574
+ }
575
+ node {
576
+ input: "63"
577
+ output: "64"
578
+ name: "Shape_41"
579
+ op_type: "Shape"
580
+ }
581
+ node {
582
+ output: "65"
583
+ name: "Constant_42"
584
+ op_type: "Constant"
585
+ attribute {
586
+ name: "value"
587
+ t {
588
+ data_type: 7
589
+ raw_data: "\000\000\000\000\000\000\000\000"
590
+ }
591
+ type: TENSOR
592
+ }
593
+ }
594
+ node {
595
+ input: "64"
596
+ input: "65"
597
+ output: "66"
598
+ name: "Gather_43"
599
+ op_type: "Gather"
600
+ attribute {
601
+ name: "axis"
602
+ i: 0
603
+ type: INT
604
+ }
605
+ }
606
+ node {
607
+ input: "63"
608
+ output: "67"
609
+ name: "Shape_44"
610
+ op_type: "Shape"
611
+ }
612
+ node {
613
+ output: "68"
614
+ name: "Constant_45"
615
+ op_type: "Constant"
616
+ attribute {
617
+ name: "value"
618
+ t {
619
+ data_type: 7
620
+ raw_data: "\001\000\000\000\000\000\000\000"
621
+ }
622
+ type: TENSOR
623
+ }
624
+ }
625
+ node {
626
+ input: "67"
627
+ input: "68"
628
+ output: "69"
629
+ name: "Gather_46"
630
+ op_type: "Gather"
631
+ attribute {
632
+ name: "axis"
633
+ i: 0
634
+ type: INT
635
+ }
636
+ }
637
+ node {
638
+ input: "63"
639
+ output: "70"
640
+ name: "Shape_47"
641
+ op_type: "Shape"
642
+ }
643
+ node {
644
+ output: "71"
645
+ name: "Constant_48"
646
+ op_type: "Constant"
647
+ attribute {
648
+ name: "value"
649
+ t {
650
+ data_type: 7
651
+ raw_data: "\002\000\000\000\000\000\000\000"
652
+ }
653
+ type: TENSOR
654
+ }
655
+ }
656
+ node {
657
+ input: "70"
658
+ input: "71"
659
+ output: "72"
660
+ name: "Gather_49"
661
+ op_type: "Gather"
662
+ attribute {
663
+ name: "axis"
664
+ i: 0
665
+ type: INT
666
+ }
667
+ }
668
+ node {
669
+ output: "73"
670
+ name: "Constant_50"
671
+ op_type: "Constant"
672
+ attribute {
673
+ name: "value"
674
+ t {
675
+ data_type: 7
676
+ raw_data: "\002\000\000\000\000\000\000\000"
677
+ }
678
+ type: TENSOR
679
+ }
680
+ }
681
+ node {
682
+ input: "69"
683
+ input: "73"
684
+ output: "74"
685
+ name: "Div_51"
686
+ op_type: "Div"
687
+ }
688
+ node {
689
+ input: "74"
690
+ output: "75"
691
+ name: "Cast_52"
692
+ op_type: "Cast"
693
+ attribute {
694
+ name: "to"
695
+ i: 7
696
+ type: INT
697
+ }
698
+ }
699
+ node {
700
+ input: "75"
701
+ output: "76"
702
+ name: "Cast_53"
703
+ op_type: "Cast"
704
+ attribute {
705
+ name: "to"
706
+ i: 7
707
+ type: INT
708
+ }
709
+ }
710
+ node {
711
+ output: "77"
712
+ name: "Constant_54"
713
+ op_type: "Constant"
714
+ attribute {
715
+ name: "value"
716
+ t {
717
+ data_type: 7
718
+ raw_data: "\002\000\000\000\000\000\000\000"
719
+ }
720
+ type: TENSOR
721
+ }
722
+ }
723
+ node {
724
+ input: "72"
725
+ input: "77"
726
+ output: "78"
727
+ name: "Mul_55"
728
+ op_type: "Mul"
729
+ }
730
+ node {
731
+ input: "66"
732
+ output: "80"
733
+ name: "Unsqueeze_56"
734
+ op_type: "Unsqueeze"
735
+ attribute {
736
+ name: "axes"
737
+ ints: 0
738
+ type: INTS
739
+ }
740
+ }
741
+ node {
742
+ input: "76"
743
+ output: "82"
744
+ name: "Unsqueeze_57"
745
+ op_type: "Unsqueeze"
746
+ attribute {
747
+ name: "axes"
748
+ ints: 0
749
+ type: INTS
750
+ }
751
+ }
752
+ node {
753
+ input: "72"
754
+ output: "83"
755
+ name: "Unsqueeze_58"
756
+ op_type: "Unsqueeze"
757
+ attribute {
758
+ name: "axes"
759
+ ints: 0
760
+ type: INTS
761
+ }
762
+ }
763
+ node {
764
+ input: "80"
765
+ input: "186"
766
+ input: "82"
767
+ input: "83"
768
+ output: "84"
769
+ name: "Concat_59"
770
+ op_type: "Concat"
771
+ attribute {
772
+ name: "axis"
773
+ i: 0
774
+ type: INT
775
+ }
776
+ }
777
+ node {
778
+ input: "63"
779
+ input: "84"
780
+ output: "85"
781
+ name: "Reshape_60"
782
+ op_type: "Reshape"
783
+ }
784
+ node {
785
+ input: "85"
786
+ output: "86"
787
+ name: "Transpose_61"
788
+ op_type: "Transpose"
789
+ attribute {
790
+ name: "perm"
791
+ ints: 0
792
+ ints: 2
793
+ ints: 3
794
+ ints: 1
795
+ type: INTS
796
+ }
797
+ }
798
+ node {
799
+ input: "66"
800
+ output: "87"
801
+ name: "Unsqueeze_62"
802
+ op_type: "Unsqueeze"
803
+ attribute {
804
+ name: "axes"
805
+ ints: 0
806
+ type: INTS
807
+ }
808
+ }
809
+ node {
810
+ input: "76"
811
+ output: "88"
812
+ name: "Unsqueeze_63"
813
+ op_type: "Unsqueeze"
814
+ attribute {
815
+ name: "axes"
816
+ ints: 0
817
+ type: INTS
818
+ }
819
+ }
820
+ node {
821
+ input: "78"
822
+ output: "89"
823
+ name: "Unsqueeze_64"
824
+ op_type: "Unsqueeze"
825
+ attribute {
826
+ name: "axes"
827
+ ints: 0
828
+ type: INTS
829
+ }
830
+ }
831
+ node {
832
+ input: "87"
833
+ input: "88"
834
+ input: "89"
835
+ output: "90"
836
+ name: "Concat_65"
837
+ op_type: "Concat"
838
+ attribute {
839
+ name: "axis"
840
+ i: 0
841
+ type: INT
842
+ }
843
+ }
844
+ node {
845
+ input: "86"
846
+ input: "90"
847
+ output: "91"
848
+ name: "Reshape_66"
849
+ op_type: "Reshape"
850
+ }
851
+ node {
852
+ input: "91"
853
+ input: "26"
854
+ output: "92"
855
+ name: "Concat_67"
856
+ op_type: "Concat"
857
+ attribute {
858
+ name: "axis"
859
+ i: 1
860
+ type: INT
861
+ }
862
+ }
863
+ node {
864
+ input: "92"
865
+ input: "up_net.1.0.weight"
866
+ input: "up_net.1.0.bias"
867
+ output: "93"
868
+ name: "Conv_68"
869
+ op_type: "Conv"
870
+ attribute {
871
+ name: "dilations"
872
+ ints: 1
873
+ type: INTS
874
+ }
875
+ attribute {
876
+ name: "group"
877
+ i: 1
878
+ type: INT
879
+ }
880
+ attribute {
881
+ name: "kernel_shape"
882
+ ints: 33
883
+ type: INTS
884
+ }
885
+ attribute {
886
+ name: "pads"
887
+ ints: 16
888
+ ints: 16
889
+ type: INTS
890
+ }
891
+ attribute {
892
+ name: "strides"
893
+ ints: 1
894
+ type: INTS
895
+ }
896
+ }
897
+ node {
898
+ input: "93"
899
+ output: "94"
900
+ name: "Relu_69"
901
+ op_type: "Relu"
902
+ }
903
+ node {
904
+ input: "94"
905
+ output: "95"
906
+ name: "Shape_70"
907
+ op_type: "Shape"
908
+ }
909
+ node {
910
+ output: "96"
911
+ name: "Constant_71"
912
+ op_type: "Constant"
913
+ attribute {
914
+ name: "value"
915
+ t {
916
+ data_type: 7
917
+ raw_data: "\000\000\000\000\000\000\000\000"
918
+ }
919
+ type: TENSOR
920
+ }
921
+ }
922
+ node {
923
+ input: "95"
924
+ input: "96"
925
+ output: "97"
926
+ name: "Gather_72"
927
+ op_type: "Gather"
928
+ attribute {
929
+ name: "axis"
930
+ i: 0
931
+ type: INT
932
+ }
933
+ }
934
+ node {
935
+ input: "94"
936
+ output: "98"
937
+ name: "Shape_73"
938
+ op_type: "Shape"
939
+ }
940
+ node {
941
+ output: "99"
942
+ name: "Constant_74"
943
+ op_type: "Constant"
944
+ attribute {
945
+ name: "value"
946
+ t {
947
+ data_type: 7
948
+ raw_data: "\001\000\000\000\000\000\000\000"
949
+ }
950
+ type: TENSOR
951
+ }
952
+ }
953
+ node {
954
+ input: "98"
955
+ input: "99"
956
+ output: "100"
957
+ name: "Gather_75"
958
+ op_type: "Gather"
959
+ attribute {
960
+ name: "axis"
961
+ i: 0
962
+ type: INT
963
+ }
964
+ }
965
+ node {
966
+ input: "94"
967
+ output: "101"
968
+ name: "Shape_76"
969
+ op_type: "Shape"
970
+ }
971
+ node {
972
+ output: "102"
973
+ name: "Constant_77"
974
+ op_type: "Constant"
975
+ attribute {
976
+ name: "value"
977
+ t {
978
+ data_type: 7
979
+ raw_data: "\002\000\000\000\000\000\000\000"
980
+ }
981
+ type: TENSOR
982
+ }
983
+ }
984
+ node {
985
+ input: "101"
986
+ input: "102"
987
+ output: "103"
988
+ name: "Gather_78"
989
+ op_type: "Gather"
990
+ attribute {
991
+ name: "axis"
992
+ i: 0
993
+ type: INT
994
+ }
995
+ }
996
+ node {
997
+ output: "104"
998
+ name: "Constant_79"
999
+ op_type: "Constant"
1000
+ attribute {
1001
+ name: "value"
1002
+ t {
1003
+ data_type: 7
1004
+ raw_data: "\002\000\000\000\000\000\000\000"
1005
+ }
1006
+ type: TENSOR
1007
+ }
1008
+ }
1009
+ node {
1010
+ input: "100"
1011
+ input: "104"
1012
+ output: "105"
1013
+ name: "Div_80"
1014
+ op_type: "Div"
1015
+ }
1016
+ node {
1017
+ input: "105"
1018
+ output: "106"
1019
+ name: "Cast_81"
1020
+ op_type: "Cast"
1021
+ attribute {
1022
+ name: "to"
1023
+ i: 7
1024
+ type: INT
1025
+ }
1026
+ }
1027
+ node {
1028
+ input: "106"
1029
+ output: "107"
1030
+ name: "Cast_82"
1031
+ op_type: "Cast"
1032
+ attribute {
1033
+ name: "to"
1034
+ i: 7
1035
+ type: INT
1036
+ }
1037
+ }
1038
+ node {
1039
+ output: "108"
1040
+ name: "Constant_83"
1041
+ op_type: "Constant"
1042
+ attribute {
1043
+ name: "value"
1044
+ t {
1045
+ data_type: 7
1046
+ raw_data: "\002\000\000\000\000\000\000\000"
1047
+ }
1048
+ type: TENSOR
1049
+ }
1050
+ }
1051
+ node {
1052
+ input: "103"
1053
+ input: "108"
1054
+ output: "109"
1055
+ name: "Mul_84"
1056
+ op_type: "Mul"
1057
+ }
1058
+ node {
1059
+ input: "97"
1060
+ output: "111"
1061
+ name: "Unsqueeze_85"
1062
+ op_type: "Unsqueeze"
1063
+ attribute {
1064
+ name: "axes"
1065
+ ints: 0
1066
+ type: INTS
1067
+ }
1068
+ }
1069
+ node {
1070
+ input: "107"
1071
+ output: "113"
1072
+ name: "Unsqueeze_86"
1073
+ op_type: "Unsqueeze"
1074
+ attribute {
1075
+ name: "axes"
1076
+ ints: 0
1077
+ type: INTS
1078
+ }
1079
+ }
1080
+ node {
1081
+ input: "103"
1082
+ output: "114"
1083
+ name: "Unsqueeze_87"
1084
+ op_type: "Unsqueeze"
1085
+ attribute {
1086
+ name: "axes"
1087
+ ints: 0
1088
+ type: INTS
1089
+ }
1090
+ }
1091
+ node {
1092
+ input: "111"
1093
+ input: "187"
1094
+ input: "113"
1095
+ input: "114"
1096
+ output: "115"
1097
+ name: "Concat_88"
1098
+ op_type: "Concat"
1099
+ attribute {
1100
+ name: "axis"
1101
+ i: 0
1102
+ type: INT
1103
+ }
1104
+ }
1105
+ node {
1106
+ input: "94"
1107
+ input: "115"
1108
+ output: "116"
1109
+ name: "Reshape_89"
1110
+ op_type: "Reshape"
1111
+ }
1112
+ node {
1113
+ input: "116"
1114
+ output: "117"
1115
+ name: "Transpose_90"
1116
+ op_type: "Transpose"
1117
+ attribute {
1118
+ name: "perm"
1119
+ ints: 0
1120
+ ints: 2
1121
+ ints: 3
1122
+ ints: 1
1123
+ type: INTS
1124
+ }
1125
+ }
1126
+ node {
1127
+ input: "97"
1128
+ output: "118"
1129
+ name: "Unsqueeze_91"
1130
+ op_type: "Unsqueeze"
1131
+ attribute {
1132
+ name: "axes"
1133
+ ints: 0
1134
+ type: INTS
1135
+ }
1136
+ }
1137
+ node {
1138
+ input: "107"
1139
+ output: "119"
1140
+ name: "Unsqueeze_92"
1141
+ op_type: "Unsqueeze"
1142
+ attribute {
1143
+ name: "axes"
1144
+ ints: 0
1145
+ type: INTS
1146
+ }
1147
+ }
1148
+ node {
1149
+ input: "109"
1150
+ output: "120"
1151
+ name: "Unsqueeze_93"
1152
+ op_type: "Unsqueeze"
1153
+ attribute {
1154
+ name: "axes"
1155
+ ints: 0
1156
+ type: INTS
1157
+ }
1158
+ }
1159
+ node {
1160
+ input: "118"
1161
+ input: "119"
1162
+ input: "120"
1163
+ output: "121"
1164
+ name: "Concat_94"
1165
+ op_type: "Concat"
1166
+ attribute {
1167
+ name: "axis"
1168
+ i: 0
1169
+ type: INT
1170
+ }
1171
+ }
1172
+ node {
1173
+ input: "117"
1174
+ input: "121"
1175
+ output: "122"
1176
+ name: "Reshape_95"
1177
+ op_type: "Reshape"
1178
+ }
1179
+ node {
1180
+ input: "122"
1181
+ input: "24"
1182
+ output: "123"
1183
+ name: "Concat_96"
1184
+ op_type: "Concat"
1185
+ attribute {
1186
+ name: "axis"
1187
+ i: 1
1188
+ type: INT
1189
+ }
1190
+ }
1191
+ node {
1192
+ input: "123"
1193
+ input: "up_net.2.0.weight"
1194
+ input: "up_net.2.0.bias"
1195
+ output: "124"
1196
+ name: "Conv_97"
1197
+ op_type: "Conv"
1198
+ attribute {
1199
+ name: "dilations"
1200
+ ints: 1
1201
+ type: INTS
1202
+ }
1203
+ attribute {
1204
+ name: "group"
1205
+ i: 1
1206
+ type: INT
1207
+ }
1208
+ attribute {
1209
+ name: "kernel_shape"
1210
+ ints: 65
1211
+ type: INTS
1212
+ }
1213
+ attribute {
1214
+ name: "pads"
1215
+ ints: 32
1216
+ ints: 32
1217
+ type: INTS
1218
+ }
1219
+ attribute {
1220
+ name: "strides"
1221
+ ints: 1
1222
+ type: INTS
1223
+ }
1224
+ }
1225
+ node {
1226
+ input: "124"
1227
+ output: "125"
1228
+ name: "Relu_98"
1229
+ op_type: "Relu"
1230
+ }
1231
+ node {
1232
+ input: "125"
1233
+ output: "126"
1234
+ name: "Shape_99"
1235
+ op_type: "Shape"
1236
+ }
1237
+ node {
1238
+ output: "127"
1239
+ name: "Constant_100"
1240
+ op_type: "Constant"
1241
+ attribute {
1242
+ name: "value"
1243
+ t {
1244
+ data_type: 7
1245
+ raw_data: "\000\000\000\000\000\000\000\000"
1246
+ }
1247
+ type: TENSOR
1248
+ }
1249
+ }
1250
+ node {
1251
+ input: "126"
1252
+ input: "127"
1253
+ output: "128"
1254
+ name: "Gather_101"
1255
+ op_type: "Gather"
1256
+ attribute {
1257
+ name: "axis"
1258
+ i: 0
1259
+ type: INT
1260
+ }
1261
+ }
1262
+ node {
1263
+ input: "125"
1264
+ output: "129"
1265
+ name: "Shape_102"
1266
+ op_type: "Shape"
1267
+ }
1268
+ node {
1269
+ output: "130"
1270
+ name: "Constant_103"
1271
+ op_type: "Constant"
1272
+ attribute {
1273
+ name: "value"
1274
+ t {
1275
+ data_type: 7
1276
+ raw_data: "\001\000\000\000\000\000\000\000"
1277
+ }
1278
+ type: TENSOR
1279
+ }
1280
+ }
1281
+ node {
1282
+ input: "129"
1283
+ input: "130"
1284
+ output: "131"
1285
+ name: "Gather_104"
1286
+ op_type: "Gather"
1287
+ attribute {
1288
+ name: "axis"
1289
+ i: 0
1290
+ type: INT
1291
+ }
1292
+ }
1293
+ node {
1294
+ input: "125"
1295
+ output: "132"
1296
+ name: "Shape_105"
1297
+ op_type: "Shape"
1298
+ }
1299
+ node {
1300
+ output: "133"
1301
+ name: "Constant_106"
1302
+ op_type: "Constant"
1303
+ attribute {
1304
+ name: "value"
1305
+ t {
1306
+ data_type: 7
1307
+ raw_data: "\002\000\000\000\000\000\000\000"
1308
+ }
1309
+ type: TENSOR
1310
+ }
1311
+ }
1312
+ node {
1313
+ input: "132"
1314
+ input: "133"
1315
+ output: "134"
1316
+ name: "Gather_107"
1317
+ op_type: "Gather"
1318
+ attribute {
1319
+ name: "axis"
1320
+ i: 0
1321
+ type: INT
1322
+ }
1323
+ }
1324
+ node {
1325
+ output: "135"
1326
+ name: "Constant_108"
1327
+ op_type: "Constant"
1328
+ attribute {
1329
+ name: "value"
1330
+ t {
1331
+ data_type: 7
1332
+ raw_data: "\002\000\000\000\000\000\000\000"
1333
+ }
1334
+ type: TENSOR
1335
+ }
1336
+ }
1337
+ node {
1338
+ input: "131"
1339
+ input: "135"
1340
+ output: "136"
1341
+ name: "Div_109"
1342
+ op_type: "Div"
1343
+ }
1344
+ node {
1345
+ input: "136"
1346
+ output: "137"
1347
+ name: "Cast_110"
1348
+ op_type: "Cast"
1349
+ attribute {
1350
+ name: "to"
1351
+ i: 7
1352
+ type: INT
1353
+ }
1354
+ }
1355
+ node {
1356
+ input: "137"
1357
+ output: "138"
1358
+ name: "Cast_111"
1359
+ op_type: "Cast"
1360
+ attribute {
1361
+ name: "to"
1362
+ i: 7
1363
+ type: INT
1364
+ }
1365
+ }
1366
+ node {
1367
+ output: "139"
1368
+ name: "Constant_112"
1369
+ op_type: "Constant"
1370
+ attribute {
1371
+ name: "value"
1372
+ t {
1373
+ data_type: 7
1374
+ raw_data: "\002\000\000\000\000\000\000\000"
1375
+ }
1376
+ type: TENSOR
1377
+ }
1378
+ }
1379
+ node {
1380
+ input: "134"
1381
+ input: "139"
1382
+ output: "140"
1383
+ name: "Mul_113"
1384
+ op_type: "Mul"
1385
+ }
1386
+ node {
1387
+ input: "128"
1388
+ output: "142"
1389
+ name: "Unsqueeze_114"
1390
+ op_type: "Unsqueeze"
1391
+ attribute {
1392
+ name: "axes"
1393
+ ints: 0
1394
+ type: INTS
1395
+ }
1396
+ }
1397
+ node {
1398
+ input: "138"
1399
+ output: "144"
1400
+ name: "Unsqueeze_115"
1401
+ op_type: "Unsqueeze"
1402
+ attribute {
1403
+ name: "axes"
1404
+ ints: 0
1405
+ type: INTS
1406
+ }
1407
+ }
1408
+ node {
1409
+ input: "134"
1410
+ output: "145"
1411
+ name: "Unsqueeze_116"
1412
+ op_type: "Unsqueeze"
1413
+ attribute {
1414
+ name: "axes"
1415
+ ints: 0
1416
+ type: INTS
1417
+ }
1418
+ }
1419
+ node {
1420
+ input: "142"
1421
+ input: "188"
1422
+ input: "144"
1423
+ input: "145"
1424
+ output: "146"
1425
+ name: "Concat_117"
1426
+ op_type: "Concat"
1427
+ attribute {
1428
+ name: "axis"
1429
+ i: 0
1430
+ type: INT
1431
+ }
1432
+ }
1433
+ node {
1434
+ input: "125"
1435
+ input: "146"
1436
+ output: "147"
1437
+ name: "Reshape_118"
1438
+ op_type: "Reshape"
1439
+ }
1440
+ node {
1441
+ input: "147"
1442
+ output: "148"
1443
+ name: "Transpose_119"
1444
+ op_type: "Transpose"
1445
+ attribute {
1446
+ name: "perm"
1447
+ ints: 0
1448
+ ints: 2
1449
+ ints: 3
1450
+ ints: 1
1451
+ type: INTS
1452
+ }
1453
+ }
1454
+ node {
1455
+ input: "128"
1456
+ output: "149"
1457
+ name: "Unsqueeze_120"
1458
+ op_type: "Unsqueeze"
1459
+ attribute {
1460
+ name: "axes"
1461
+ ints: 0
1462
+ type: INTS
1463
+ }
1464
+ }
1465
+ node {
1466
+ input: "138"
1467
+ output: "150"
1468
+ name: "Unsqueeze_121"
1469
+ op_type: "Unsqueeze"
1470
+ attribute {
1471
+ name: "axes"
1472
+ ints: 0
1473
+ type: INTS
1474
+ }
1475
+ }
1476
+ node {
1477
+ input: "140"
1478
+ output: "151"
1479
+ name: "Unsqueeze_122"
1480
+ op_type: "Unsqueeze"
1481
+ attribute {
1482
+ name: "axes"
1483
+ ints: 0
1484
+ type: INTS
1485
+ }
1486
+ }
1487
+ node {
1488
+ input: "149"
1489
+ input: "150"
1490
+ input: "151"
1491
+ output: "152"
1492
+ name: "Concat_123"
1493
+ op_type: "Concat"
1494
+ attribute {
1495
+ name: "axis"
1496
+ i: 0
1497
+ type: INT
1498
+ }
1499
+ }
1500
+ node {
1501
+ input: "148"
1502
+ input: "152"
1503
+ output: "153"
1504
+ name: "Reshape_124"
1505
+ op_type: "Reshape"
1506
+ }
1507
+ node {
1508
+ input: "153"
1509
+ input: "22"
1510
+ output: "154"
1511
+ name: "Concat_125"
1512
+ op_type: "Concat"
1513
+ attribute {
1514
+ name: "axis"
1515
+ i: 1
1516
+ type: INT
1517
+ }
1518
+ }
1519
+ node {
1520
+ input: "154"
1521
+ input: "up_net.3.0.weight"
1522
+ input: "up_net.3.0.bias"
1523
+ output: "155"
1524
+ name: "Conv_126"
1525
+ op_type: "Conv"
1526
+ attribute {
1527
+ name: "dilations"
1528
+ ints: 1
1529
+ type: INTS
1530
+ }
1531
+ attribute {
1532
+ name: "group"
1533
+ i: 1
1534
+ type: INT
1535
+ }
1536
+ attribute {
1537
+ name: "kernel_shape"
1538
+ ints: 9
1539
+ type: INTS
1540
+ }
1541
+ attribute {
1542
+ name: "pads"
1543
+ ints: 4
1544
+ ints: 4
1545
+ type: INTS
1546
+ }
1547
+ attribute {
1548
+ name: "strides"
1549
+ ints: 1
1550
+ type: INTS
1551
+ }
1552
+ }
1553
+ node {
1554
+ input: "155"
1555
+ output: "156"
1556
+ name: "Shape_127"
1557
+ op_type: "Shape"
1558
+ }
1559
+ node {
1560
+ output: "157"
1561
+ name: "Constant_128"
1562
+ op_type: "Constant"
1563
+ attribute {
1564
+ name: "value"
1565
+ t {
1566
+ data_type: 7
1567
+ raw_data: "\000\000\000\000\000\000\000\000"
1568
+ }
1569
+ type: TENSOR
1570
+ }
1571
+ }
1572
+ node {
1573
+ input: "156"
1574
+ input: "157"
1575
+ output: "158"
1576
+ name: "Gather_129"
1577
+ op_type: "Gather"
1578
+ attribute {
1579
+ name: "axis"
1580
+ i: 0
1581
+ type: INT
1582
+ }
1583
+ }
1584
+ node {
1585
+ input: "155"
1586
+ output: "159"
1587
+ name: "Shape_130"
1588
+ op_type: "Shape"
1589
+ }
1590
+ node {
1591
+ output: "160"
1592
+ name: "Constant_131"
1593
+ op_type: "Constant"
1594
+ attribute {
1595
+ name: "value"
1596
+ t {
1597
+ data_type: 7
1598
+ raw_data: "\001\000\000\000\000\000\000\000"
1599
+ }
1600
+ type: TENSOR
1601
+ }
1602
+ }
1603
+ node {
1604
+ input: "159"
1605
+ input: "160"
1606
+ output: "161"
1607
+ name: "Gather_132"
1608
+ op_type: "Gather"
1609
+ attribute {
1610
+ name: "axis"
1611
+ i: 0
1612
+ type: INT
1613
+ }
1614
+ }
1615
+ node {
1616
+ input: "155"
1617
+ output: "162"
1618
+ name: "Shape_133"
1619
+ op_type: "Shape"
1620
+ }
1621
+ node {
1622
+ output: "163"
1623
+ name: "Constant_134"
1624
+ op_type: "Constant"
1625
+ attribute {
1626
+ name: "value"
1627
+ t {
1628
+ data_type: 7
1629
+ raw_data: "\002\000\000\000\000\000\000\000"
1630
+ }
1631
+ type: TENSOR
1632
+ }
1633
+ }
1634
+ node {
1635
+ input: "162"
1636
+ input: "163"
1637
+ output: "164"
1638
+ name: "Gather_135"
1639
+ op_type: "Gather"
1640
+ attribute {
1641
+ name: "axis"
1642
+ i: 0
1643
+ type: INT
1644
+ }
1645
+ }
1646
+ node {
1647
+ output: "165"
1648
+ name: "Constant_136"
1649
+ op_type: "Constant"
1650
+ attribute {
1651
+ name: "value"
1652
+ t {
1653
+ data_type: 7
1654
+ raw_data: "\002\000\000\000\000\000\000\000"
1655
+ }
1656
+ type: TENSOR
1657
+ }
1658
+ }
1659
+ node {
1660
+ input: "161"
1661
+ input: "165"
1662
+ output: "166"
1663
+ name: "Div_137"
1664
+ op_type: "Div"
1665
+ }
1666
+ node {
1667
+ input: "166"
1668
+ output: "167"
1669
+ name: "Cast_138"
1670
+ op_type: "Cast"
1671
+ attribute {
1672
+ name: "to"
1673
+ i: 7
1674
+ type: INT
1675
+ }
1676
+ }
1677
+ node {
1678
+ input: "167"
1679
+ output: "168"
1680
+ name: "Cast_139"
1681
+ op_type: "Cast"
1682
+ attribute {
1683
+ name: "to"
1684
+ i: 7
1685
+ type: INT
1686
+ }
1687
+ }
1688
+ node {
1689
+ output: "169"
1690
+ name: "Constant_140"
1691
+ op_type: "Constant"
1692
+ attribute {
1693
+ name: "value"
1694
+ t {
1695
+ data_type: 7
1696
+ raw_data: "\002\000\000\000\000\000\000\000"
1697
+ }
1698
+ type: TENSOR
1699
+ }
1700
+ }
1701
+ node {
1702
+ input: "164"
1703
+ input: "169"
1704
+ output: "170"
1705
+ name: "Mul_141"
1706
+ op_type: "Mul"
1707
+ }
1708
+ node {
1709
+ input: "158"
1710
+ output: "172"
1711
+ name: "Unsqueeze_142"
1712
+ op_type: "Unsqueeze"
1713
+ attribute {
1714
+ name: "axes"
1715
+ ints: 0
1716
+ type: INTS
1717
+ }
1718
+ }
1719
+ node {
1720
+ input: "168"
1721
+ output: "174"
1722
+ name: "Unsqueeze_143"
1723
+ op_type: "Unsqueeze"
1724
+ attribute {
1725
+ name: "axes"
1726
+ ints: 0
1727
+ type: INTS
1728
+ }
1729
+ }
1730
+ node {
1731
+ input: "164"
1732
+ output: "175"
1733
+ name: "Unsqueeze_144"
1734
+ op_type: "Unsqueeze"
1735
+ attribute {
1736
+ name: "axes"
1737
+ ints: 0
1738
+ type: INTS
1739
+ }
1740
+ }
1741
+ node {
1742
+ input: "172"
1743
+ input: "189"
1744
+ input: "174"
1745
+ input: "175"
1746
+ output: "176"
1747
+ name: "Concat_145"
1748
+ op_type: "Concat"
1749
+ attribute {
1750
+ name: "axis"
1751
+ i: 0
1752
+ type: INT
1753
+ }
1754
+ }
1755
+ node {
1756
+ input: "155"
1757
+ input: "176"
1758
+ output: "177"
1759
+ name: "Reshape_146"
1760
+ op_type: "Reshape"
1761
+ }
1762
+ node {
1763
+ input: "177"
1764
+ output: "178"
1765
+ name: "Transpose_147"
1766
+ op_type: "Transpose"
1767
+ attribute {
1768
+ name: "perm"
1769
+ ints: 0
1770
+ ints: 2
1771
+ ints: 3
1772
+ ints: 1
1773
+ type: INTS
1774
+ }
1775
+ }
1776
+ node {
1777
+ input: "158"
1778
+ output: "179"
1779
+ name: "Unsqueeze_148"
1780
+ op_type: "Unsqueeze"
1781
+ attribute {
1782
+ name: "axes"
1783
+ ints: 0
1784
+ type: INTS
1785
+ }
1786
+ }
1787
+ node {
1788
+ input: "168"
1789
+ output: "180"
1790
+ name: "Unsqueeze_149"
1791
+ op_type: "Unsqueeze"
1792
+ attribute {
1793
+ name: "axes"
1794
+ ints: 0
1795
+ type: INTS
1796
+ }
1797
+ }
1798
+ node {
1799
+ input: "170"
1800
+ output: "181"
1801
+ name: "Unsqueeze_150"
1802
+ op_type: "Unsqueeze"
1803
+ attribute {
1804
+ name: "axes"
1805
+ ints: 0
1806
+ type: INTS
1807
+ }
1808
+ }
1809
+ node {
1810
+ input: "179"
1811
+ input: "180"
1812
+ input: "181"
1813
+ output: "182"
1814
+ name: "Concat_151"
1815
+ op_type: "Concat"
1816
+ attribute {
1817
+ name: "axis"
1818
+ i: 0
1819
+ type: INT
1820
+ }
1821
+ }
1822
+ node {
1823
+ input: "178"
1824
+ input: "182"
1825
+ output: "183"
1826
+ name: "Reshape_152"
1827
+ op_type: "Reshape"
1828
+ }
1829
+ node {
1830
+ input: "183"
1831
+ input: "x"
1832
+ output: "y"
1833
+ name: "Add_153"
1834
+ op_type: "Add"
1835
+ }
1836
+ initializer {
1837
+ dims: 1
1838
+ data_type: 7
1839
+ name: "185"
1840
+ raw_data: "\002\000\000\000\000\000\000\000"
1841
+ }
1842
+ initializer {
1843
+ dims: 1
1844
+ data_type: 7
1845
+ name: "186"
1846
+ raw_data: "\002\000\000\000\000\000\000\000"
1847
+ }
1848
+ initializer {
1849
+ dims: 1
1850
+ data_type: 7
1851
+ name: "187"
1852
+ raw_data: "\002\000\000\000\000\000\000\000"
1853
+ }
1854
+ initializer {
1855
+ dims: 1
1856
+ data_type: 7
1857
+ name: "188"
1858
+ raw_data: "\002\000\000\000\000\000\000\000"
1859
+ }
1860
+ initializer {
1861
+ dims: 1
1862
+ data_type: 7
1863
+ name: "189"
1864
+ raw_data: "\002\000\000\000\000\000\000\000"
1865
+ }
1866
+ initializer {
1867
+ dims: 512
1868
+ data_type: 1
1869
+ name: "bottleneck.0.bias"
1870
+ }
1871
+ initializer {
1872
+ dims: 512
1873
+ dims: 512
1874
+ dims: 9
1875
+ data_type: 1
1876
+ name: "bottleneck.0.weight"
1877
+ }
1878
+ initializer {
1879
+ dims: 1024
1880
+ data_type: 1
1881
+ name: "bottleneck.3.bias"
1882
+ }
1883
+ initializer {
1884
+ dims: 1024
1885
+ dims: 512
1886
+ dims: 9
1887
+ data_type: 1
1888
+ name: "bottleneck.3.weight"
1889
+ }
1890
+ initializer {
1891
+ dims: 128
1892
+ data_type: 1
1893
+ name: "down_net.0.0.bias"
1894
+ }
1895
+ initializer {
1896
+ dims: 128
1897
+ dims: 2
1898
+ dims: 65
1899
+ data_type: 1
1900
+ name: "down_net.0.0.weight"
1901
+ }
1902
+ initializer {
1903
+ dims: 256
1904
+ data_type: 1
1905
+ name: "down_net.1.0.bias"
1906
+ }
1907
+ initializer {
1908
+ dims: 256
1909
+ dims: 128
1910
+ dims: 33
1911
+ data_type: 1
1912
+ name: "down_net.1.0.weight"
1913
+ }
1914
+ initializer {
1915
+ dims: 512
1916
+ data_type: 1
1917
+ name: "down_net.2.0.bias"
1918
+ }
1919
+ initializer {
1920
+ dims: 512
1921
+ dims: 256
1922
+ dims: 17
1923
+ data_type: 1
1924
+ name: "down_net.2.0.weight"
1925
+ }
1926
+ initializer {
1927
+ dims: 512
1928
+ data_type: 1
1929
+ name: "down_net.3.0.bias"
1930
+ }
1931
+ initializer {
1932
+ dims: 512
1933
+ dims: 512
1934
+ dims: 9
1935
+ data_type: 1
1936
+ name: "down_net.3.0.weight"
1937
+ }
1938
+ initializer {
1939
+ dims: 1024
1940
+ data_type: 1
1941
+ name: "up_net.0.0.bias"
1942
+ }
1943
+ initializer {
1944
+ dims: 1024
1945
+ dims: 1024
1946
+ dims: 17
1947
+ data_type: 1
1948
+ name: "up_net.0.0.weight"
1949
+ }
1950
+ initializer {
1951
+ dims: 512
1952
+ data_type: 1
1953
+ name: "up_net.1.0.bias"
1954
+ }
1955
+ initializer {
1956
+ dims: 512
1957
+ dims: 1024
1958
+ dims: 33
1959
+ data_type: 1
1960
+ name: "up_net.1.0.weight"
1961
+ }
1962
+ initializer {
1963
+ dims: 256
1964
+ data_type: 1
1965
+ name: "up_net.2.0.bias"
1966
+ }
1967
+ initializer {
1968
+ dims: 256
1969
+ dims: 512
1970
+ dims: 65
1971
+ data_type: 1
1972
+ name: "up_net.2.0.weight"
1973
+ }
1974
+ initializer {
1975
+ dims: 4
1976
+ data_type: 1
1977
+ name: "up_net.3.0.bias"
1978
+ raw_data: "\245\002g\274\024\365\247\272\310\301\210\274\032\367k\274"
1979
+ }
1980
+ initializer {
1981
+ dims: 4
1982
+ dims: 256
1983
+ dims: 9
1984
+ data_type: 1
1985
+ name: "up_net.3.0.weight"
1986
+ }
1987
+ input {
1988
+ name: "x"
1989
+ type {
1990
+ tensor_type {
1991
+ elem_type: 1
1992
+ shape {
1993
+ dim {
1994
+ dim_param: "batch_size"
1995
+ }
1996
+ dim {
1997
+ dim_value: 2
1998
+ }
1999
+ dim {
2000
+ dim_value: 8192
2001
+ }
2002
+ }
2003
+ }
2004
+ }
2005
+ }
2006
+ output {
2007
+ name: "y"
2008
+ type {
2009
+ tensor_type {
2010
+ elem_type: 1
2011
+ shape {
2012
+ dim {
2013
+ dim_param: "batch_size"
2014
+ }
2015
+ dim {
2016
+ dim_value: 2
2017
+ }
2018
+ dim {
2019
+ dim_value: 8192
2020
+ }
2021
+ }
2022
+ }
2023
+ }
2024
+ }
2025
+ }
2026
+ opset_import {
2027
+ version: 11
2028
+ }