Spaces:
Sleeping
Sleeping
first commit
Browse files- JointModel.png +0 -0
- LICENSE +661 -0
- data_loader.py +269 -0
- dataset_statistic.png +0 -0
- early_stopping.py +64 -0
- gradio_demo.py +250 -0
- main.py +139 -0
- predict.py +232 -0
- predict.sh +3 -0
- requirements.txt +11 -0
- run_jointBERT-CRF_PhoBERTencoder.sh +23 -0
- run_jointBERT-CRF_XLM-Rencoder.sh +23 -0
- run_jointIDSF_PhoBERTencoder.sh +30 -0
- run_jointIDSF_XLM-Rencoder.sh +30 -0
- trainer.py +300 -0
- utils.py +115 -0
JointModel.png
ADDED
![]() |
LICENSE
ADDED
@@ -0,0 +1,661 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
GNU AFFERO GENERAL PUBLIC LICENSE
|
2 |
+
Version 3, 19 November 2007
|
3 |
+
|
4 |
+
Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
|
5 |
+
Everyone is permitted to copy and distribute verbatim copies
|
6 |
+
of this license document, but changing it is not allowed.
|
7 |
+
|
8 |
+
Preamble
|
9 |
+
|
10 |
+
The GNU Affero General Public License is a free, copyleft license for
|
11 |
+
software and other kinds of works, specifically designed to ensure
|
12 |
+
cooperation with the community in the case of network server software.
|
13 |
+
|
14 |
+
The licenses for most software and other practical works are designed
|
15 |
+
to take away your freedom to share and change the works. By contrast,
|
16 |
+
our General Public Licenses are intended to guarantee your freedom to
|
17 |
+
share and change all versions of a program--to make sure it remains free
|
18 |
+
software for all its users.
|
19 |
+
|
20 |
+
When we speak of free software, we are referring to freedom, not
|
21 |
+
price. Our General Public Licenses are designed to make sure that you
|
22 |
+
have the freedom to distribute copies of free software (and charge for
|
23 |
+
them if you wish), that you receive source code or can get it if you
|
24 |
+
want it, that you can change the software or use pieces of it in new
|
25 |
+
free programs, and that you know you can do these things.
|
26 |
+
|
27 |
+
Developers that use our General Public Licenses protect your rights
|
28 |
+
with two steps: (1) assert copyright on the software, and (2) offer
|
29 |
+
you this License which gives you legal permission to copy, distribute
|
30 |
+
and/or modify the software.
|
31 |
+
|
32 |
+
A secondary benefit of defending all users' freedom is that
|
33 |
+
improvements made in alternate versions of the program, if they
|
34 |
+
receive widespread use, become available for other developers to
|
35 |
+
incorporate. Many developers of free software are heartened and
|
36 |
+
encouraged by the resulting cooperation. However, in the case of
|
37 |
+
software used on network servers, this result may fail to come about.
|
38 |
+
The GNU General Public License permits making a modified version and
|
39 |
+
letting the public access it on a server without ever releasing its
|
40 |
+
source code to the public.
|
41 |
+
|
42 |
+
The GNU Affero General Public License is designed specifically to
|
43 |
+
ensure that, in such cases, the modified source code becomes available
|
44 |
+
to the community. It requires the operator of a network server to
|
45 |
+
provide the source code of the modified version running there to the
|
46 |
+
users of that server. Therefore, public use of a modified version, on
|
47 |
+
a publicly accessible server, gives the public access to the source
|
48 |
+
code of the modified version.
|
49 |
+
|
50 |
+
An older license, called the Affero General Public License and
|
51 |
+
published by Affero, was designed to accomplish similar goals. This is
|
52 |
+
a different license, not a version of the Affero GPL, but Affero has
|
53 |
+
released a new version of the Affero GPL which permits relicensing under
|
54 |
+
this license.
|
55 |
+
|
56 |
+
The precise terms and conditions for copying, distribution and
|
57 |
+
modification follow.
|
58 |
+
|
59 |
+
TERMS AND CONDITIONS
|
60 |
+
|
61 |
+
0. Definitions.
|
62 |
+
|
63 |
+
"This License" refers to version 3 of the GNU Affero General Public License.
|
64 |
+
|
65 |
+
"Copyright" also means copyright-like laws that apply to other kinds of
|
66 |
+
works, such as semiconductor masks.
|
67 |
+
|
68 |
+
"The Program" refers to any copyrightable work licensed under this
|
69 |
+
License. Each licensee is addressed as "you". "Licensees" and
|
70 |
+
"recipients" may be individuals or organizations.
|
71 |
+
|
72 |
+
To "modify" a work means to copy from or adapt all or part of the work
|
73 |
+
in a fashion requiring copyright permission, other than the making of an
|
74 |
+
exact copy. The resulting work is called a "modified version" of the
|
75 |
+
earlier work or a work "based on" the earlier work.
|
76 |
+
|
77 |
+
A "covered work" means either the unmodified Program or a work based
|
78 |
+
on the Program.
|
79 |
+
|
80 |
+
To "propagate" a work means to do anything with it that, without
|
81 |
+
permission, would make you directly or secondarily liable for
|
82 |
+
infringement under applicable copyright law, except executing it on a
|
83 |
+
computer or modifying a private copy. Propagation includes copying,
|
84 |
+
distribution (with or without modification), making available to the
|
85 |
+
public, and in some countries other activities as well.
|
86 |
+
|
87 |
+
To "convey" a work means any kind of propagation that enables other
|
88 |
+
parties to make or receive copies. Mere interaction with a user through
|
89 |
+
a computer network, with no transfer of a copy, is not conveying.
|
90 |
+
|
91 |
+
An interactive user interface displays "Appropriate Legal Notices"
|
92 |
+
to the extent that it includes a convenient and prominently visible
|
93 |
+
feature that (1) displays an appropriate copyright notice, and (2)
|
94 |
+
tells the user that there is no warranty for the work (except to the
|
95 |
+
extent that warranties are provided), that licensees may convey the
|
96 |
+
work under this License, and how to view a copy of this License. If
|
97 |
+
the interface presents a list of user commands or options, such as a
|
98 |
+
menu, a prominent item in the list meets this criterion.
|
99 |
+
|
100 |
+
1. Source Code.
|
101 |
+
|
102 |
+
The "source code" for a work means the preferred form of the work
|
103 |
+
for making modifications to it. "Object code" means any non-source
|
104 |
+
form of a work.
|
105 |
+
|
106 |
+
A "Standard Interface" means an interface that either is an official
|
107 |
+
standard defined by a recognized standards body, or, in the case of
|
108 |
+
interfaces specified for a particular programming language, one that
|
109 |
+
is widely used among developers working in that language.
|
110 |
+
|
111 |
+
The "System Libraries" of an executable work include anything, other
|
112 |
+
than the work as a whole, that (a) is included in the normal form of
|
113 |
+
packaging a Major Component, but which is not part of that Major
|
114 |
+
Component, and (b) serves only to enable use of the work with that
|
115 |
+
Major Component, or to implement a Standard Interface for which an
|
116 |
+
implementation is available to the public in source code form. A
|
117 |
+
"Major Component", in this context, means a major essential component
|
118 |
+
(kernel, window system, and so on) of the specific operating system
|
119 |
+
(if any) on which the executable work runs, or a compiler used to
|
120 |
+
produce the work, or an object code interpreter used to run it.
|
121 |
+
|
122 |
+
The "Corresponding Source" for a work in object code form means all
|
123 |
+
the source code needed to generate, install, and (for an executable
|
124 |
+
work) run the object code and to modify the work, including scripts to
|
125 |
+
control those activities. However, it does not include the work's
|
126 |
+
System Libraries, or general-purpose tools or generally available free
|
127 |
+
programs which are used unmodified in performing those activities but
|
128 |
+
which are not part of the work. For example, Corresponding Source
|
129 |
+
includes interface definition files associated with source files for
|
130 |
+
the work, and the source code for shared libraries and dynamically
|
131 |
+
linked subprograms that the work is specifically designed to require,
|
132 |
+
such as by intimate data communication or control flow between those
|
133 |
+
subprograms and other parts of the work.
|
134 |
+
|
135 |
+
The Corresponding Source need not include anything that users
|
136 |
+
can regenerate automatically from other parts of the Corresponding
|
137 |
+
Source.
|
138 |
+
|
139 |
+
The Corresponding Source for a work in source code form is that
|
140 |
+
same work.
|
141 |
+
|
142 |
+
2. Basic Permissions.
|
143 |
+
|
144 |
+
All rights granted under this License are granted for the term of
|
145 |
+
copyright on the Program, and are irrevocable provided the stated
|
146 |
+
conditions are met. This License explicitly affirms your unlimited
|
147 |
+
permission to run the unmodified Program. The output from running a
|
148 |
+
covered work is covered by this License only if the output, given its
|
149 |
+
content, constitutes a covered work. This License acknowledges your
|
150 |
+
rights of fair use or other equivalent, as provided by copyright law.
|
151 |
+
|
152 |
+
You may make, run and propagate covered works that you do not
|
153 |
+
convey, without conditions so long as your license otherwise remains
|
154 |
+
in force. You may convey covered works to others for the sole purpose
|
155 |
+
of having them make modifications exclusively for you, or provide you
|
156 |
+
with facilities for running those works, provided that you comply with
|
157 |
+
the terms of this License in conveying all material for which you do
|
158 |
+
not control copyright. Those thus making or running the covered works
|
159 |
+
for you must do so exclusively on your behalf, under your direction
|
160 |
+
and control, on terms that prohibit them from making any copies of
|
161 |
+
your copyrighted material outside their relationship with you.
|
162 |
+
|
163 |
+
Conveying under any other circumstances is permitted solely under
|
164 |
+
the conditions stated below. Sublicensing is not allowed; section 10
|
165 |
+
makes it unnecessary.
|
166 |
+
|
167 |
+
3. Protecting Users' Legal Rights From Anti-Circumvention Law.
|
168 |
+
|
169 |
+
No covered work shall be deemed part of an effective technological
|
170 |
+
measure under any applicable law fulfilling obligations under article
|
171 |
+
11 of the WIPO copyright treaty adopted on 20 December 1996, or
|
172 |
+
similar laws prohibiting or restricting circumvention of such
|
173 |
+
measures.
|
174 |
+
|
175 |
+
When you convey a covered work, you waive any legal power to forbid
|
176 |
+
circumvention of technological measures to the extent such circumvention
|
177 |
+
is effected by exercising rights under this License with respect to
|
178 |
+
the covered work, and you disclaim any intention to limit operation or
|
179 |
+
modification of the work as a means of enforcing, against the work's
|
180 |
+
users, your or third parties' legal rights to forbid circumvention of
|
181 |
+
technological measures.
|
182 |
+
|
183 |
+
4. Conveying Verbatim Copies.
|
184 |
+
|
185 |
+
You may convey verbatim copies of the Program's source code as you
|
186 |
+
receive it, in any medium, provided that you conspicuously and
|
187 |
+
appropriately publish on each copy an appropriate copyright notice;
|
188 |
+
keep intact all notices stating that this License and any
|
189 |
+
non-permissive terms added in accord with section 7 apply to the code;
|
190 |
+
keep intact all notices of the absence of any warranty; and give all
|
191 |
+
recipients a copy of this License along with the Program.
|
192 |
+
|
193 |
+
You may charge any price or no price for each copy that you convey,
|
194 |
+
and you may offer support or warranty protection for a fee.
|
195 |
+
|
196 |
+
5. Conveying Modified Source Versions.
|
197 |
+
|
198 |
+
You may convey a work based on the Program, or the modifications to
|
199 |
+
produce it from the Program, in the form of source code under the
|
200 |
+
terms of section 4, provided that you also meet all of these conditions:
|
201 |
+
|
202 |
+
a) The work must carry prominent notices stating that you modified
|
203 |
+
it, and giving a relevant date.
|
204 |
+
|
205 |
+
b) The work must carry prominent notices stating that it is
|
206 |
+
released under this License and any conditions added under section
|
207 |
+
7. This requirement modifies the requirement in section 4 to
|
208 |
+
"keep intact all notices".
|
209 |
+
|
210 |
+
c) You must license the entire work, as a whole, under this
|
211 |
+
License to anyone who comes into possession of a copy. This
|
212 |
+
License will therefore apply, along with any applicable section 7
|
213 |
+
additional terms, to the whole of the work, and all its parts,
|
214 |
+
regardless of how they are packaged. This License gives no
|
215 |
+
permission to license the work in any other way, but it does not
|
216 |
+
invalidate such permission if you have separately received it.
|
217 |
+
|
218 |
+
d) If the work has interactive user interfaces, each must display
|
219 |
+
Appropriate Legal Notices; however, if the Program has interactive
|
220 |
+
interfaces that do not display Appropriate Legal Notices, your
|
221 |
+
work need not make them do so.
|
222 |
+
|
223 |
+
A compilation of a covered work with other separate and independent
|
224 |
+
works, which are not by their nature extensions of the covered work,
|
225 |
+
and which are not combined with it such as to form a larger program,
|
226 |
+
in or on a volume of a storage or distribution medium, is called an
|
227 |
+
"aggregate" if the compilation and its resulting copyright are not
|
228 |
+
used to limit the access or legal rights of the compilation's users
|
229 |
+
beyond what the individual works permit. Inclusion of a covered work
|
230 |
+
in an aggregate does not cause this License to apply to the other
|
231 |
+
parts of the aggregate.
|
232 |
+
|
233 |
+
6. Conveying Non-Source Forms.
|
234 |
+
|
235 |
+
You may convey a covered work in object code form under the terms
|
236 |
+
of sections 4 and 5, provided that you also convey the
|
237 |
+
machine-readable Corresponding Source under the terms of this License,
|
238 |
+
in one of these ways:
|
239 |
+
|
240 |
+
a) Convey the object code in, or embodied in, a physical product
|
241 |
+
(including a physical distribution medium), accompanied by the
|
242 |
+
Corresponding Source fixed on a durable physical medium
|
243 |
+
customarily used for software interchange.
|
244 |
+
|
245 |
+
b) Convey the object code in, or embodied in, a physical product
|
246 |
+
(including a physical distribution medium), accompanied by a
|
247 |
+
written offer, valid for at least three years and valid for as
|
248 |
+
long as you offer spare parts or customer support for that product
|
249 |
+
model, to give anyone who possesses the object code either (1) a
|
250 |
+
copy of the Corresponding Source for all the software in the
|
251 |
+
product that is covered by this License, on a durable physical
|
252 |
+
medium customarily used for software interchange, for a price no
|
253 |
+
more than your reasonable cost of physically performing this
|
254 |
+
conveying of source, or (2) access to copy the
|
255 |
+
Corresponding Source from a network server at no charge.
|
256 |
+
|
257 |
+
c) Convey individual copies of the object code with a copy of the
|
258 |
+
written offer to provide the Corresponding Source. This
|
259 |
+
alternative is allowed only occasionally and noncommercially, and
|
260 |
+
only if you received the object code with such an offer, in accord
|
261 |
+
with subsection 6b.
|
262 |
+
|
263 |
+
d) Convey the object code by offering access from a designated
|
264 |
+
place (gratis or for a charge), and offer equivalent access to the
|
265 |
+
Corresponding Source in the same way through the same place at no
|
266 |
+
further charge. You need not require recipients to copy the
|
267 |
+
Corresponding Source along with the object code. If the place to
|
268 |
+
copy the object code is a network server, the Corresponding Source
|
269 |
+
may be on a different server (operated by you or a third party)
|
270 |
+
that supports equivalent copying facilities, provided you maintain
|
271 |
+
clear directions next to the object code saying where to find the
|
272 |
+
Corresponding Source. Regardless of what server hosts the
|
273 |
+
Corresponding Source, you remain obligated to ensure that it is
|
274 |
+
available for as long as needed to satisfy these requirements.
|
275 |
+
|
276 |
+
e) Convey the object code using peer-to-peer transmission, provided
|
277 |
+
you inform other peers where the object code and Corresponding
|
278 |
+
Source of the work are being offered to the general public at no
|
279 |
+
charge under subsection 6d.
|
280 |
+
|
281 |
+
A separable portion of the object code, whose source code is excluded
|
282 |
+
from the Corresponding Source as a System Library, need not be
|
283 |
+
included in conveying the object code work.
|
284 |
+
|
285 |
+
A "User Product" is either (1) a "consumer product", which means any
|
286 |
+
tangible personal property which is normally used for personal, family,
|
287 |
+
or household purposes, or (2) anything designed or sold for incorporation
|
288 |
+
into a dwelling. In determining whether a product is a consumer product,
|
289 |
+
doubtful cases shall be resolved in favor of coverage. For a particular
|
290 |
+
product received by a particular user, "normally used" refers to a
|
291 |
+
typical or common use of that class of product, regardless of the status
|
292 |
+
of the particular user or of the way in which the particular user
|
293 |
+
actually uses, or expects or is expected to use, the product. A product
|
294 |
+
is a consumer product regardless of whether the product has substantial
|
295 |
+
commercial, industrial or non-consumer uses, unless such uses represent
|
296 |
+
the only significant mode of use of the product.
|
297 |
+
|
298 |
+
"Installation Information" for a User Product means any methods,
|
299 |
+
procedures, authorization keys, or other information required to install
|
300 |
+
and execute modified versions of a covered work in that User Product from
|
301 |
+
a modified version of its Corresponding Source. The information must
|
302 |
+
suffice to ensure that the continued functioning of the modified object
|
303 |
+
code is in no case prevented or interfered with solely because
|
304 |
+
modification has been made.
|
305 |
+
|
306 |
+
If you convey an object code work under this section in, or with, or
|
307 |
+
specifically for use in, a User Product, and the conveying occurs as
|
308 |
+
part of a transaction in which the right of possession and use of the
|
309 |
+
User Product is transferred to the recipient in perpetuity or for a
|
310 |
+
fixed term (regardless of how the transaction is characterized), the
|
311 |
+
Corresponding Source conveyed under this section must be accompanied
|
312 |
+
by the Installation Information. But this requirement does not apply
|
313 |
+
if neither you nor any third party retains the ability to install
|
314 |
+
modified object code on the User Product (for example, the work has
|
315 |
+
been installed in ROM).
|
316 |
+
|
317 |
+
The requirement to provide Installation Information does not include a
|
318 |
+
requirement to continue to provide support service, warranty, or updates
|
319 |
+
for a work that has been modified or installed by the recipient, or for
|
320 |
+
the User Product in which it has been modified or installed. Access to a
|
321 |
+
network may be denied when the modification itself materially and
|
322 |
+
adversely affects the operation of the network or violates the rules and
|
323 |
+
protocols for communication across the network.
|
324 |
+
|
325 |
+
Corresponding Source conveyed, and Installation Information provided,
|
326 |
+
in accord with this section must be in a format that is publicly
|
327 |
+
documented (and with an implementation available to the public in
|
328 |
+
source code form), and must require no special password or key for
|
329 |
+
unpacking, reading or copying.
|
330 |
+
|
331 |
+
7. Additional Terms.
|
332 |
+
|
333 |
+
"Additional permissions" are terms that supplement the terms of this
|
334 |
+
License by making exceptions from one or more of its conditions.
|
335 |
+
Additional permissions that are applicable to the entire Program shall
|
336 |
+
be treated as though they were included in this License, to the extent
|
337 |
+
that they are valid under applicable law. If additional permissions
|
338 |
+
apply only to part of the Program, that part may be used separately
|
339 |
+
under those permissions, but the entire Program remains governed by
|
340 |
+
this License without regard to the additional permissions.
|
341 |
+
|
342 |
+
When you convey a copy of a covered work, you may at your option
|
343 |
+
remove any additional permissions from that copy, or from any part of
|
344 |
+
it. (Additional permissions may be written to require their own
|
345 |
+
removal in certain cases when you modify the work.) You may place
|
346 |
+
additional permissions on material, added by you to a covered work,
|
347 |
+
for which you have or can give appropriate copyright permission.
|
348 |
+
|
349 |
+
Notwithstanding any other provision of this License, for material you
|
350 |
+
add to a covered work, you may (if authorized by the copyright holders of
|
351 |
+
that material) supplement the terms of this License with terms:
|
352 |
+
|
353 |
+
a) Disclaiming warranty or limiting liability differently from the
|
354 |
+
terms of sections 15 and 16 of this License; or
|
355 |
+
|
356 |
+
b) Requiring preservation of specified reasonable legal notices or
|
357 |
+
author attributions in that material or in the Appropriate Legal
|
358 |
+
Notices displayed by works containing it; or
|
359 |
+
|
360 |
+
c) Prohibiting misrepresentation of the origin of that material, or
|
361 |
+
requiring that modified versions of such material be marked in
|
362 |
+
reasonable ways as different from the original version; or
|
363 |
+
|
364 |
+
d) Limiting the use for publicity purposes of names of licensors or
|
365 |
+
authors of the material; or
|
366 |
+
|
367 |
+
e) Declining to grant rights under trademark law for use of some
|
368 |
+
trade names, trademarks, or service marks; or
|
369 |
+
|
370 |
+
f) Requiring indemnification of licensors and authors of that
|
371 |
+
material by anyone who conveys the material (or modified versions of
|
372 |
+
it) with contractual assumptions of liability to the recipient, for
|
373 |
+
any liability that these contractual assumptions directly impose on
|
374 |
+
those licensors and authors.
|
375 |
+
|
376 |
+
All other non-permissive additional terms are considered "further
|
377 |
+
restrictions" within the meaning of section 10. If the Program as you
|
378 |
+
received it, or any part of it, contains a notice stating that it is
|
379 |
+
governed by this License along with a term that is a further
|
380 |
+
restriction, you may remove that term. If a license document contains
|
381 |
+
a further restriction but permits relicensing or conveying under this
|
382 |
+
License, you may add to a covered work material governed by the terms
|
383 |
+
of that license document, provided that the further restriction does
|
384 |
+
not survive such relicensing or conveying.
|
385 |
+
|
386 |
+
If you add terms to a covered work in accord with this section, you
|
387 |
+
must place, in the relevant source files, a statement of the
|
388 |
+
additional terms that apply to those files, or a notice indicating
|
389 |
+
where to find the applicable terms.
|
390 |
+
|
391 |
+
Additional terms, permissive or non-permissive, may be stated in the
|
392 |
+
form of a separately written license, or stated as exceptions;
|
393 |
+
the above requirements apply either way.
|
394 |
+
|
395 |
+
8. Termination.
|
396 |
+
|
397 |
+
You may not propagate or modify a covered work except as expressly
|
398 |
+
provided under this License. Any attempt otherwise to propagate or
|
399 |
+
modify it is void, and will automatically terminate your rights under
|
400 |
+
this License (including any patent licenses granted under the third
|
401 |
+
paragraph of section 11).
|
402 |
+
|
403 |
+
However, if you cease all violation of this License, then your
|
404 |
+
license from a particular copyright holder is reinstated (a)
|
405 |
+
provisionally, unless and until the copyright holder explicitly and
|
406 |
+
finally terminates your license, and (b) permanently, if the copyright
|
407 |
+
holder fails to notify you of the violation by some reasonable means
|
408 |
+
prior to 60 days after the cessation.
|
409 |
+
|
410 |
+
Moreover, your license from a particular copyright holder is
|
411 |
+
reinstated permanently if the copyright holder notifies you of the
|
412 |
+
violation by some reasonable means, this is the first time you have
|
413 |
+
received notice of violation of this License (for any work) from that
|
414 |
+
copyright holder, and you cure the violation prior to 30 days after
|
415 |
+
your receipt of the notice.
|
416 |
+
|
417 |
+
Termination of your rights under this section does not terminate the
|
418 |
+
licenses of parties who have received copies or rights from you under
|
419 |
+
this License. If your rights have been terminated and not permanently
|
420 |
+
reinstated, you do not qualify to receive new licenses for the same
|
421 |
+
material under section 10.
|
422 |
+
|
423 |
+
9. Acceptance Not Required for Having Copies.
|
424 |
+
|
425 |
+
You are not required to accept this License in order to receive or
|
426 |
+
run a copy of the Program. Ancillary propagation of a covered work
|
427 |
+
occurring solely as a consequence of using peer-to-peer transmission
|
428 |
+
to receive a copy likewise does not require acceptance. However,
|
429 |
+
nothing other than this License grants you permission to propagate or
|
430 |
+
modify any covered work. These actions infringe copyright if you do
|
431 |
+
not accept this License. Therefore, by modifying or propagating a
|
432 |
+
covered work, you indicate your acceptance of this License to do so.
|
433 |
+
|
434 |
+
10. Automatic Licensing of Downstream Recipients.
|
435 |
+
|
436 |
+
Each time you convey a covered work, the recipient automatically
|
437 |
+
receives a license from the original licensors, to run, modify and
|
438 |
+
propagate that work, subject to this License. You are not responsible
|
439 |
+
for enforcing compliance by third parties with this License.
|
440 |
+
|
441 |
+
An "entity transaction" is a transaction transferring control of an
|
442 |
+
organization, or substantially all assets of one, or subdividing an
|
443 |
+
organization, or merging organizations. If propagation of a covered
|
444 |
+
work results from an entity transaction, each party to that
|
445 |
+
transaction who receives a copy of the work also receives whatever
|
446 |
+
licenses to the work the party's predecessor in interest had or could
|
447 |
+
give under the previous paragraph, plus a right to possession of the
|
448 |
+
Corresponding Source of the work from the predecessor in interest, if
|
449 |
+
the predecessor has it or can get it with reasonable efforts.
|
450 |
+
|
451 |
+
You may not impose any further restrictions on the exercise of the
|
452 |
+
rights granted or affirmed under this License. For example, you may
|
453 |
+
not impose a license fee, royalty, or other charge for exercise of
|
454 |
+
rights granted under this License, and you may not initiate litigation
|
455 |
+
(including a cross-claim or counterclaim in a lawsuit) alleging that
|
456 |
+
any patent claim is infringed by making, using, selling, offering for
|
457 |
+
sale, or importing the Program or any portion of it.
|
458 |
+
|
459 |
+
11. Patents.
|
460 |
+
|
461 |
+
A "contributor" is a copyright holder who authorizes use under this
|
462 |
+
License of the Program or a work on which the Program is based. The
|
463 |
+
work thus licensed is called the contributor's "contributor version".
|
464 |
+
|
465 |
+
A contributor's "essential patent claims" are all patent claims
|
466 |
+
owned or controlled by the contributor, whether already acquired or
|
467 |
+
hereafter acquired, that would be infringed by some manner, permitted
|
468 |
+
by this License, of making, using, or selling its contributor version,
|
469 |
+
but do not include claims that would be infringed only as a
|
470 |
+
consequence of further modification of the contributor version. For
|
471 |
+
purposes of this definition, "control" includes the right to grant
|
472 |
+
patent sublicenses in a manner consistent with the requirements of
|
473 |
+
this License.
|
474 |
+
|
475 |
+
Each contributor grants you a non-exclusive, worldwide, royalty-free
|
476 |
+
patent license under the contributor's essential patent claims, to
|
477 |
+
make, use, sell, offer for sale, import and otherwise run, modify and
|
478 |
+
propagate the contents of its contributor version.
|
479 |
+
|
480 |
+
In the following three paragraphs, a "patent license" is any express
|
481 |
+
agreement or commitment, however denominated, not to enforce a patent
|
482 |
+
(such as an express permission to practice a patent or covenant not to
|
483 |
+
sue for patent infringement). To "grant" such a patent license to a
|
484 |
+
party means to make such an agreement or commitment not to enforce a
|
485 |
+
patent against the party.
|
486 |
+
|
487 |
+
If you convey a covered work, knowingly relying on a patent license,
|
488 |
+
and the Corresponding Source of the work is not available for anyone
|
489 |
+
to copy, free of charge and under the terms of this License, through a
|
490 |
+
publicly available network server or other readily accessible means,
|
491 |
+
then you must either (1) cause the Corresponding Source to be so
|
492 |
+
available, or (2) arrange to deprive yourself of the benefit of the
|
493 |
+
patent license for this particular work, or (3) arrange, in a manner
|
494 |
+
consistent with the requirements of this License, to extend the patent
|
495 |
+
license to downstream recipients. "Knowingly relying" means you have
|
496 |
+
actual knowledge that, but for the patent license, your conveying the
|
497 |
+
covered work in a country, or your recipient's use of the covered work
|
498 |
+
in a country, would infringe one or more identifiable patents in that
|
499 |
+
country that you have reason to believe are valid.
|
500 |
+
|
501 |
+
If, pursuant to or in connection with a single transaction or
|
502 |
+
arrangement, you convey, or propagate by procuring conveyance of, a
|
503 |
+
covered work, and grant a patent license to some of the parties
|
504 |
+
receiving the covered work authorizing them to use, propagate, modify
|
505 |
+
or convey a specific copy of the covered work, then the patent license
|
506 |
+
you grant is automatically extended to all recipients of the covered
|
507 |
+
work and works based on it.
|
508 |
+
|
509 |
+
A patent license is "discriminatory" if it does not include within
|
510 |
+
the scope of its coverage, prohibits the exercise of, or is
|
511 |
+
conditioned on the non-exercise of one or more of the rights that are
|
512 |
+
specifically granted under this License. You may not convey a covered
|
513 |
+
work if you are a party to an arrangement with a third party that is
|
514 |
+
in the business of distributing software, under which you make payment
|
515 |
+
to the third party based on the extent of your activity of conveying
|
516 |
+
the work, and under which the third party grants, to any of the
|
517 |
+
parties who would receive the covered work from you, a discriminatory
|
518 |
+
patent license (a) in connection with copies of the covered work
|
519 |
+
conveyed by you (or copies made from those copies), or (b) primarily
|
520 |
+
for and in connection with specific products or compilations that
|
521 |
+
contain the covered work, unless you entered into that arrangement,
|
522 |
+
or that patent license was granted, prior to 28 March 2007.
|
523 |
+
|
524 |
+
Nothing in this License shall be construed as excluding or limiting
|
525 |
+
any implied license or other defenses to infringement that may
|
526 |
+
otherwise be available to you under applicable patent law.
|
527 |
+
|
528 |
+
12. No Surrender of Others' Freedom.
|
529 |
+
|
530 |
+
If conditions are imposed on you (whether by court order, agreement or
|
531 |
+
otherwise) that contradict the conditions of this License, they do not
|
532 |
+
excuse you from the conditions of this License. If you cannot convey a
|
533 |
+
covered work so as to satisfy simultaneously your obligations under this
|
534 |
+
License and any other pertinent obligations, then as a consequence you may
|
535 |
+
not convey it at all. For example, if you agree to terms that obligate you
|
536 |
+
to collect a royalty for further conveying from those to whom you convey
|
537 |
+
the Program, the only way you could satisfy both those terms and this
|
538 |
+
License would be to refrain entirely from conveying the Program.
|
539 |
+
|
540 |
+
13. Remote Network Interaction; Use with the GNU General Public License.
|
541 |
+
|
542 |
+
Notwithstanding any other provision of this License, if you modify the
|
543 |
+
Program, your modified version must prominently offer all users
|
544 |
+
interacting with it remotely through a computer network (if your version
|
545 |
+
supports such interaction) an opportunity to receive the Corresponding
|
546 |
+
Source of your version by providing access to the Corresponding Source
|
547 |
+
from a network server at no charge, through some standard or customary
|
548 |
+
means of facilitating copying of software. This Corresponding Source
|
549 |
+
shall include the Corresponding Source for any work covered by version 3
|
550 |
+
of the GNU General Public License that is incorporated pursuant to the
|
551 |
+
following paragraph.
|
552 |
+
|
553 |
+
Notwithstanding any other provision of this License, you have
|
554 |
+
permission to link or combine any covered work with a work licensed
|
555 |
+
under version 3 of the GNU General Public License into a single
|
556 |
+
combined work, and to convey the resulting work. The terms of this
|
557 |
+
License will continue to apply to the part which is the covered work,
|
558 |
+
but the work with which it is combined will remain governed by version
|
559 |
+
3 of the GNU General Public License.
|
560 |
+
|
561 |
+
14. Revised Versions of this License.
|
562 |
+
|
563 |
+
The Free Software Foundation may publish revised and/or new versions of
|
564 |
+
the GNU Affero General Public License from time to time. Such new versions
|
565 |
+
will be similar in spirit to the present version, but may differ in detail to
|
566 |
+
address new problems or concerns.
|
567 |
+
|
568 |
+
Each version is given a distinguishing version number. If the
|
569 |
+
Program specifies that a certain numbered version of the GNU Affero General
|
570 |
+
Public License "or any later version" applies to it, you have the
|
571 |
+
option of following the terms and conditions either of that numbered
|
572 |
+
version or of any later version published by the Free Software
|
573 |
+
Foundation. If the Program does not specify a version number of the
|
574 |
+
GNU Affero General Public License, you may choose any version ever published
|
575 |
+
by the Free Software Foundation.
|
576 |
+
|
577 |
+
If the Program specifies that a proxy can decide which future
|
578 |
+
versions of the GNU Affero General Public License can be used, that proxy's
|
579 |
+
public statement of acceptance of a version permanently authorizes you
|
580 |
+
to choose that version for the Program.
|
581 |
+
|
582 |
+
Later license versions may give you additional or different
|
583 |
+
permissions. However, no additional obligations are imposed on any
|
584 |
+
author or copyright holder as a result of your choosing to follow a
|
585 |
+
later version.
|
586 |
+
|
587 |
+
15. Disclaimer of Warranty.
|
588 |
+
|
589 |
+
THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
|
590 |
+
APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
|
591 |
+
HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
|
592 |
+
OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
|
593 |
+
THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
594 |
+
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
|
595 |
+
IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
|
596 |
+
ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
|
597 |
+
|
598 |
+
16. Limitation of Liability.
|
599 |
+
|
600 |
+
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
|
601 |
+
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
|
602 |
+
THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
|
603 |
+
GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
|
604 |
+
USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
|
605 |
+
DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
|
606 |
+
PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
|
607 |
+
EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
|
608 |
+
SUCH DAMAGES.
|
609 |
+
|
610 |
+
17. Interpretation of Sections 15 and 16.
|
611 |
+
|
612 |
+
If the disclaimer of warranty and limitation of liability provided
|
613 |
+
above cannot be given local legal effect according to their terms,
|
614 |
+
reviewing courts shall apply local law that most closely approximates
|
615 |
+
an absolute waiver of all civil liability in connection with the
|
616 |
+
Program, unless a warranty or assumption of liability accompanies a
|
617 |
+
copy of the Program in return for a fee.
|
618 |
+
|
619 |
+
END OF TERMS AND CONDITIONS
|
620 |
+
|
621 |
+
How to Apply These Terms to Your New Programs
|
622 |
+
|
623 |
+
If you develop a new program, and you want it to be of the greatest
|
624 |
+
possible use to the public, the best way to achieve this is to make it
|
625 |
+
free software which everyone can redistribute and change under these terms.
|
626 |
+
|
627 |
+
To do so, attach the following notices to the program. It is safest
|
628 |
+
to attach them to the start of each source file to most effectively
|
629 |
+
state the exclusion of warranty; and each file should have at least
|
630 |
+
the "copyright" line and a pointer to where the full notice is found.
|
631 |
+
|
632 |
+
<one line to give the program's name and a brief idea of what it does.>
|
633 |
+
Copyright (C) <year> <name of author>
|
634 |
+
|
635 |
+
This program is free software: you can redistribute it and/or modify
|
636 |
+
it under the terms of the GNU Affero General Public License as published
|
637 |
+
by the Free Software Foundation, either version 3 of the License, or
|
638 |
+
(at your option) any later version.
|
639 |
+
|
640 |
+
This program is distributed in the hope that it will be useful,
|
641 |
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
642 |
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
643 |
+
GNU Affero General Public License for more details.
|
644 |
+
|
645 |
+
You should have received a copy of the GNU Affero General Public License
|
646 |
+
along with this program. If not, see <https://www.gnu.org/licenses/>.
|
647 |
+
|
648 |
+
Also add information on how to contact you by electronic and paper mail.
|
649 |
+
|
650 |
+
If your software can interact with users remotely through a computer
|
651 |
+
network, you should also make sure that it provides a way for users to
|
652 |
+
get its source. For example, if your program is a web application, its
|
653 |
+
interface could display a "Source" link that leads users to an archive
|
654 |
+
of the code. There are many ways you could offer source, and different
|
655 |
+
solutions will be better for different programs; see section 13 for the
|
656 |
+
specific requirements.
|
657 |
+
|
658 |
+
You should also get your employer (if you work as a programmer) or school,
|
659 |
+
if any, to sign a "copyright disclaimer" for the program, if necessary.
|
660 |
+
For more information on this, and how to apply and follow the GNU AGPL, see
|
661 |
+
<https://www.gnu.org/licenses/>.
|
data_loader.py
ADDED
@@ -0,0 +1,269 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import copy
|
2 |
+
import json
|
3 |
+
import logging
|
4 |
+
import os
|
5 |
+
|
6 |
+
import torch
|
7 |
+
from torch.utils.data import TensorDataset
|
8 |
+
from utils import get_intent_labels, get_slot_labels
|
9 |
+
|
10 |
+
|
11 |
+
logger = logging.getLogger(__name__)
|
12 |
+
|
13 |
+
|
14 |
+
class InputExample(object):
|
15 |
+
"""
|
16 |
+
A single training/test example for simple sequence classification.
|
17 |
+
|
18 |
+
Args:
|
19 |
+
guid: Unique id for the example.
|
20 |
+
words: list. The words of the sequence.
|
21 |
+
intent_label: (Optional) string. The intent label of the example.
|
22 |
+
slot_labels: (Optional) list. The slot labels of the example.
|
23 |
+
"""
|
24 |
+
|
25 |
+
def __init__(self, guid, words, intent_label=None, slot_labels=None):
|
26 |
+
self.guid = guid
|
27 |
+
self.words = words
|
28 |
+
self.intent_label = intent_label
|
29 |
+
self.slot_labels = slot_labels
|
30 |
+
|
31 |
+
def __repr__(self):
|
32 |
+
return str(self.to_json_string())
|
33 |
+
|
34 |
+
def to_dict(self):
|
35 |
+
"""Serializes this instance to a Python dictionary."""
|
36 |
+
output = copy.deepcopy(self.__dict__)
|
37 |
+
return output
|
38 |
+
|
39 |
+
def to_json_string(self):
|
40 |
+
"""Serializes this instance to a JSON string."""
|
41 |
+
return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n"
|
42 |
+
|
43 |
+
|
44 |
+
class InputFeatures(object):
|
45 |
+
"""A single set of features of data."""
|
46 |
+
|
47 |
+
def __init__(self, input_ids, attention_mask, token_type_ids, intent_label_id, slot_labels_ids):
|
48 |
+
self.input_ids = input_ids
|
49 |
+
self.attention_mask = attention_mask
|
50 |
+
self.token_type_ids = token_type_ids
|
51 |
+
self.intent_label_id = intent_label_id
|
52 |
+
self.slot_labels_ids = slot_labels_ids
|
53 |
+
|
54 |
+
def __repr__(self):
|
55 |
+
return str(self.to_json_string())
|
56 |
+
|
57 |
+
def to_dict(self):
|
58 |
+
"""Serializes this instance to a Python dictionary."""
|
59 |
+
output = copy.deepcopy(self.__dict__)
|
60 |
+
return output
|
61 |
+
|
62 |
+
def to_json_string(self):
|
63 |
+
"""Serializes this instance to a JSON string."""
|
64 |
+
return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n"
|
65 |
+
|
66 |
+
|
67 |
+
class JointProcessor(object):
|
68 |
+
"""Processor for the JointBERT data set """
|
69 |
+
|
70 |
+
def __init__(self, args):
|
71 |
+
self.args = args
|
72 |
+
self.intent_labels = get_intent_labels(args)
|
73 |
+
self.slot_labels = get_slot_labels(args)
|
74 |
+
|
75 |
+
self.input_text_file = "seq.in"
|
76 |
+
self.intent_label_file = "label"
|
77 |
+
self.slot_labels_file = "seq.out"
|
78 |
+
|
79 |
+
@classmethod
|
80 |
+
def _read_file(cls, input_file, quotechar=None):
|
81 |
+
"""Reads a tab separated value file."""
|
82 |
+
with open(input_file, "r", encoding="utf-8") as f:
|
83 |
+
lines = []
|
84 |
+
for line in f:
|
85 |
+
lines.append(line.strip())
|
86 |
+
return lines
|
87 |
+
|
88 |
+
def _create_examples(self, texts, intents, slots, set_type):
|
89 |
+
"""Creates examples for the training and dev sets."""
|
90 |
+
examples = []
|
91 |
+
for i, (text, intent, slot) in enumerate(zip(texts, intents, slots)):
|
92 |
+
guid = "%s-%s" % (set_type, i)
|
93 |
+
# 1. input_text
|
94 |
+
words = text.split() # Some are spaced twice
|
95 |
+
# 2. intent
|
96 |
+
intent_label = (
|
97 |
+
self.intent_labels.index(intent) if intent in self.intent_labels else self.intent_labels.index("UNK")
|
98 |
+
)
|
99 |
+
# 3. slot
|
100 |
+
slot_labels = []
|
101 |
+
for s in slot.split():
|
102 |
+
slot_labels.append(
|
103 |
+
self.slot_labels.index(s) if s in self.slot_labels else self.slot_labels.index("UNK")
|
104 |
+
)
|
105 |
+
|
106 |
+
assert len(words) == len(slot_labels)
|
107 |
+
examples.append(InputExample(guid=guid, words=words, intent_label=intent_label, slot_labels=slot_labels))
|
108 |
+
return examples
|
109 |
+
|
110 |
+
def get_examples(self, mode):
|
111 |
+
"""
|
112 |
+
Args:
|
113 |
+
mode: train, dev, test
|
114 |
+
"""
|
115 |
+
data_path = os.path.join(self.args.data_dir, self.args.token_level, mode)
|
116 |
+
logger.info("LOOKING AT {}".format(data_path))
|
117 |
+
return self._create_examples(
|
118 |
+
texts=self._read_file(os.path.join(data_path, self.input_text_file)),
|
119 |
+
intents=self._read_file(os.path.join(data_path, self.intent_label_file)),
|
120 |
+
slots=self._read_file(os.path.join(data_path, self.slot_labels_file)),
|
121 |
+
set_type=mode,
|
122 |
+
)
|
123 |
+
|
124 |
+
|
125 |
+
processors = {"syllable-level": JointProcessor, "word-level": JointProcessor}
|
126 |
+
|
127 |
+
|
128 |
+
def convert_examples_to_features(
|
129 |
+
examples,
|
130 |
+
max_seq_len,
|
131 |
+
tokenizer,
|
132 |
+
pad_token_label_id=-100,
|
133 |
+
cls_token_segment_id=0,
|
134 |
+
pad_token_segment_id=0,
|
135 |
+
sequence_a_segment_id=0,
|
136 |
+
mask_padding_with_zero=True,
|
137 |
+
):
|
138 |
+
# Setting based on the current model type
|
139 |
+
cls_token = tokenizer.cls_token
|
140 |
+
sep_token = tokenizer.sep_token
|
141 |
+
unk_token = tokenizer.unk_token
|
142 |
+
pad_token_id = tokenizer.pad_token_id
|
143 |
+
|
144 |
+
features = []
|
145 |
+
for (ex_index, example) in enumerate(examples):
|
146 |
+
if ex_index % 5000 == 0:
|
147 |
+
logger.info("Writing example %d of %d" % (ex_index, len(examples)))
|
148 |
+
|
149 |
+
# Tokenize word by word (for NER)
|
150 |
+
tokens = []
|
151 |
+
slot_labels_ids = []
|
152 |
+
for word, slot_label in zip(example.words, example.slot_labels):
|
153 |
+
word_tokens = tokenizer.tokenize(word)
|
154 |
+
if not word_tokens:
|
155 |
+
word_tokens = [unk_token] # For handling the bad-encoded word
|
156 |
+
tokens.extend(word_tokens)
|
157 |
+
# Use the real label id for the first token of the word, and padding ids for the remaining tokens
|
158 |
+
slot_labels_ids.extend([int(slot_label)] + [pad_token_label_id] * (len(word_tokens) - 1))
|
159 |
+
|
160 |
+
# Account for [CLS] and [SEP]
|
161 |
+
special_tokens_count = 2
|
162 |
+
if len(tokens) > max_seq_len - special_tokens_count:
|
163 |
+
tokens = tokens[: (max_seq_len - special_tokens_count)]
|
164 |
+
slot_labels_ids = slot_labels_ids[: (max_seq_len - special_tokens_count)]
|
165 |
+
|
166 |
+
# Add [SEP] token
|
167 |
+
tokens += [sep_token]
|
168 |
+
slot_labels_ids += [pad_token_label_id]
|
169 |
+
token_type_ids = [sequence_a_segment_id] * len(tokens)
|
170 |
+
|
171 |
+
# Add [CLS] token
|
172 |
+
tokens = [cls_token] + tokens
|
173 |
+
slot_labels_ids = [pad_token_label_id] + slot_labels_ids
|
174 |
+
token_type_ids = [cls_token_segment_id] + token_type_ids
|
175 |
+
|
176 |
+
input_ids = tokenizer.convert_tokens_to_ids(tokens)
|
177 |
+
|
178 |
+
# The mask has 1 for real tokens and 0 for padding tokens. Only real
|
179 |
+
# tokens are attended to.
|
180 |
+
attention_mask = [1 if mask_padding_with_zero else 0] * len(input_ids)
|
181 |
+
|
182 |
+
# Zero-pad up to the sequence length.
|
183 |
+
padding_length = max_seq_len - len(input_ids)
|
184 |
+
input_ids = input_ids + ([pad_token_id] * padding_length)
|
185 |
+
attention_mask = attention_mask + ([0 if mask_padding_with_zero else 1] * padding_length)
|
186 |
+
token_type_ids = token_type_ids + ([pad_token_segment_id] * padding_length)
|
187 |
+
slot_labels_ids = slot_labels_ids + ([pad_token_label_id] * padding_length)
|
188 |
+
|
189 |
+
assert len(input_ids) == max_seq_len, "Error with input length {} vs {}".format(len(input_ids), max_seq_len)
|
190 |
+
assert len(attention_mask) == max_seq_len, "Error with attention mask length {} vs {}".format(
|
191 |
+
len(attention_mask), max_seq_len
|
192 |
+
)
|
193 |
+
assert len(token_type_ids) == max_seq_len, "Error with token type length {} vs {}".format(
|
194 |
+
len(token_type_ids), max_seq_len
|
195 |
+
)
|
196 |
+
assert len(slot_labels_ids) == max_seq_len, "Error with slot labels length {} vs {}".format(
|
197 |
+
len(slot_labels_ids), max_seq_len
|
198 |
+
)
|
199 |
+
|
200 |
+
intent_label_id = int(example.intent_label)
|
201 |
+
|
202 |
+
if ex_index < 5:
|
203 |
+
logger.info("*** Example ***")
|
204 |
+
logger.info("guid: %s" % example.guid)
|
205 |
+
logger.info("tokens: %s" % " ".join([str(x) for x in tokens]))
|
206 |
+
logger.info("input_ids: %s" % " ".join([str(x) for x in input_ids]))
|
207 |
+
logger.info("attention_mask: %s" % " ".join([str(x) for x in attention_mask]))
|
208 |
+
logger.info("token_type_ids: %s" % " ".join([str(x) for x in token_type_ids]))
|
209 |
+
logger.info("intent_label: %s (id = %d)" % (example.intent_label, intent_label_id))
|
210 |
+
logger.info("slot_labels: %s" % " ".join([str(x) for x in slot_labels_ids]))
|
211 |
+
|
212 |
+
features.append(
|
213 |
+
InputFeatures(
|
214 |
+
input_ids=input_ids,
|
215 |
+
attention_mask=attention_mask,
|
216 |
+
token_type_ids=token_type_ids,
|
217 |
+
intent_label_id=intent_label_id,
|
218 |
+
slot_labels_ids=slot_labels_ids,
|
219 |
+
)
|
220 |
+
)
|
221 |
+
|
222 |
+
return features
|
223 |
+
|
224 |
+
|
225 |
+
def load_and_cache_examples(args, tokenizer, mode):
|
226 |
+
processor = processors[args.token_level](args)
|
227 |
+
|
228 |
+
# Load data features from cache or dataset file
|
229 |
+
cached_features_file = os.path.join(
|
230 |
+
args.data_dir,
|
231 |
+
"cached_{}_{}_{}_{}".format(
|
232 |
+
mode, args.token_level, list(filter(None, args.model_name_or_path.split("/"))).pop(), args.max_seq_len
|
233 |
+
),
|
234 |
+
)
|
235 |
+
|
236 |
+
if os.path.exists(cached_features_file):
|
237 |
+
logger.info("Loading features from cached file %s", cached_features_file)
|
238 |
+
features = torch.load(cached_features_file)
|
239 |
+
else:
|
240 |
+
# Load data features from dataset file
|
241 |
+
logger.info("Creating features from dataset file at %s", args.data_dir)
|
242 |
+
if mode == "train":
|
243 |
+
examples = processor.get_examples("train")
|
244 |
+
elif mode == "dev":
|
245 |
+
examples = processor.get_examples("dev")
|
246 |
+
elif mode == "test":
|
247 |
+
examples = processor.get_examples("test")
|
248 |
+
else:
|
249 |
+
raise Exception("For mode, Only train, dev, test is available")
|
250 |
+
|
251 |
+
# Use cross entropy ignore index as padding label id so that only real label ids contribute to the loss later
|
252 |
+
pad_token_label_id = args.ignore_index
|
253 |
+
features = convert_examples_to_features(
|
254 |
+
examples, args.max_seq_len, tokenizer, pad_token_label_id=pad_token_label_id
|
255 |
+
)
|
256 |
+
logger.info("Saving features into cached file %s", cached_features_file)
|
257 |
+
torch.save(features, cached_features_file)
|
258 |
+
|
259 |
+
# Convert to Tensors and build dataset
|
260 |
+
all_input_ids = torch.tensor([f.input_ids for f in features], dtype=torch.long)
|
261 |
+
all_attention_mask = torch.tensor([f.attention_mask for f in features], dtype=torch.long)
|
262 |
+
all_token_type_ids = torch.tensor([f.token_type_ids for f in features], dtype=torch.long)
|
263 |
+
all_intent_label_ids = torch.tensor([f.intent_label_id for f in features], dtype=torch.long)
|
264 |
+
all_slot_labels_ids = torch.tensor([f.slot_labels_ids for f in features], dtype=torch.long)
|
265 |
+
|
266 |
+
dataset = TensorDataset(
|
267 |
+
all_input_ids, all_attention_mask, all_token_type_ids, all_intent_label_ids, all_slot_labels_ids
|
268 |
+
)
|
269 |
+
return dataset
|
dataset_statistic.png
ADDED
![]() |
early_stopping.py
ADDED
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
|
3 |
+
import numpy as np
|
4 |
+
import torch
|
5 |
+
|
6 |
+
|
7 |
+
class EarlyStopping:
|
8 |
+
"""Early stops the training if validation loss doesn't improve after a given patience."""
|
9 |
+
|
10 |
+
def __init__(self, patience=7, verbose=False):
|
11 |
+
"""
|
12 |
+
Args:
|
13 |
+
patience (int): How long to wait after last time validation loss improved.
|
14 |
+
Default: 7
|
15 |
+
verbose (bool): If True, prints a message for each validation loss improvement.
|
16 |
+
Default: False
|
17 |
+
"""
|
18 |
+
self.patience = patience
|
19 |
+
self.verbose = verbose
|
20 |
+
self.counter = 0
|
21 |
+
self.best_score = None
|
22 |
+
self.early_stop = False
|
23 |
+
self.val_loss_min = np.Inf
|
24 |
+
|
25 |
+
def __call__(self, val_loss, model, args):
|
26 |
+
if args.tuning_metric == "loss":
|
27 |
+
score = -val_loss
|
28 |
+
else:
|
29 |
+
score = val_loss
|
30 |
+
if self.best_score is None:
|
31 |
+
self.best_score = score
|
32 |
+
self.save_checkpoint(val_loss, model, args)
|
33 |
+
elif score < self.best_score:
|
34 |
+
self.counter += 1
|
35 |
+
print(f"EarlyStopping counter: {self.counter} out of {self.patience}")
|
36 |
+
if self.counter >= self.patience:
|
37 |
+
self.early_stop = True
|
38 |
+
else:
|
39 |
+
self.best_score = score
|
40 |
+
self.save_checkpoint(val_loss, model, args)
|
41 |
+
self.counter = 0
|
42 |
+
|
43 |
+
def save_checkpoint(self, val_loss, model, args):
|
44 |
+
"""Saves model when validation loss decreases or accuracy/f1 increases."""
|
45 |
+
if self.verbose:
|
46 |
+
if args.tuning_metric == "loss":
|
47 |
+
print(f"Validation loss decreased ({self.val_loss_min:.6f} --> {val_loss:.6f}). Saving model ...")
|
48 |
+
else:
|
49 |
+
print(
|
50 |
+
f"{args.tuning_metric} increased ({self.val_loss_min:.6f} --> {val_loss:.6f}). Saving model ..."
|
51 |
+
)
|
52 |
+
model.save_pretrained(args.model_dir)
|
53 |
+
torch.save(args, os.path.join(args.model_dir, "training_args.bin"))
|
54 |
+
self.val_loss_min = val_loss
|
55 |
+
|
56 |
+
# # Save model checkpoint (Overwrite)
|
57 |
+
# if not os.path.exists(self.args.model_dir):
|
58 |
+
# os.makedirs(self.args.model_dir)
|
59 |
+
# model_to_save = self.model.module if hasattr(self.model, 'module') else self.model
|
60 |
+
# model_to_save.save_pretrained(self.args.model_dir)
|
61 |
+
|
62 |
+
# # Save training arguments together with the trained model
|
63 |
+
# torch.save(self.args, os.path.join(self.args.model_dir, 'training_args.bin'))
|
64 |
+
# logger.info("Saving model checkpoint to %s", self.args.model_dir)
|
gradio_demo.py
ADDED
@@ -0,0 +1,250 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
|
3 |
+
import argparse
|
4 |
+
import logging
|
5 |
+
import os
|
6 |
+
|
7 |
+
import numpy as np
|
8 |
+
import torch
|
9 |
+
from torch.utils.data import DataLoader, SequentialSampler, TensorDataset
|
10 |
+
from tqdm import tqdm
|
11 |
+
from utils import MODEL_CLASSES, get_intent_labels, get_slot_labels, init_logger, load_tokenizer
|
12 |
+
|
13 |
+
|
14 |
+
logger = logging.getLogger(__name__)
|
15 |
+
|
16 |
+
|
17 |
+
def get_device(pred_config):
|
18 |
+
return "cuda" if torch.cuda.is_available() and not pred_config.no_cuda else "cpu"
|
19 |
+
|
20 |
+
|
21 |
+
def get_args(pred_config):
|
22 |
+
args = torch.load(os.path.join(pred_config.model_dir, "training_args.bin"))
|
23 |
+
|
24 |
+
args.model_dir = pred_config.model_dir
|
25 |
+
args.data_dir = 'PhoATIS'
|
26 |
+
|
27 |
+
return args
|
28 |
+
|
29 |
+
|
30 |
+
def load_model(pred_config, args, device):
|
31 |
+
# Check whether model exists
|
32 |
+
if not os.path.exists(pred_config.model_dir):
|
33 |
+
raise Exception("Model doesn't exists! Train first!")
|
34 |
+
|
35 |
+
try:
|
36 |
+
model = MODEL_CLASSES[args.model_type][1].from_pretrained(
|
37 |
+
args.model_dir, args=args, intent_label_lst=get_intent_labels(args), slot_label_lst=get_slot_labels(args)
|
38 |
+
)
|
39 |
+
model.to(device)
|
40 |
+
model.eval()
|
41 |
+
logger.info("***** Model Loaded *****")
|
42 |
+
except Exception:
|
43 |
+
raise Exception("Some model files might be missing...")
|
44 |
+
|
45 |
+
return model
|
46 |
+
|
47 |
+
def convert_input_file_to_tensor_dataset(
|
48 |
+
lines,
|
49 |
+
pred_config,
|
50 |
+
args,
|
51 |
+
tokenizer,
|
52 |
+
pad_token_label_id,
|
53 |
+
cls_token_segment_id=0,
|
54 |
+
pad_token_segment_id=0,
|
55 |
+
sequence_a_segment_id=0,
|
56 |
+
mask_padding_with_zero=True,
|
57 |
+
):
|
58 |
+
# Setting based on the current model type
|
59 |
+
cls_token = tokenizer.cls_token
|
60 |
+
sep_token = tokenizer.sep_token
|
61 |
+
unk_token = tokenizer.unk_token
|
62 |
+
pad_token_id = tokenizer.pad_token_id
|
63 |
+
|
64 |
+
all_input_ids = []
|
65 |
+
all_attention_mask = []
|
66 |
+
all_token_type_ids = []
|
67 |
+
all_slot_label_mask = []
|
68 |
+
|
69 |
+
for words in lines:
|
70 |
+
tokens = []
|
71 |
+
slot_label_mask = []
|
72 |
+
for word in words:
|
73 |
+
word_tokens = tokenizer.tokenize(word)
|
74 |
+
if not word_tokens:
|
75 |
+
word_tokens = [unk_token] # For handling the bad-encoded word
|
76 |
+
tokens.extend(word_tokens)
|
77 |
+
# Use the real label id for the first token of the word, and padding ids for the remaining tokens
|
78 |
+
slot_label_mask.extend([pad_token_label_id + 1] + [pad_token_label_id] * (len(word_tokens) - 1))
|
79 |
+
|
80 |
+
# Account for [CLS] and [SEP]
|
81 |
+
special_tokens_count = 2
|
82 |
+
if len(tokens) > args.max_seq_len - special_tokens_count:
|
83 |
+
tokens = tokens[: (args.max_seq_len - special_tokens_count)]
|
84 |
+
slot_label_mask = slot_label_mask[: (args.max_seq_len - special_tokens_count)]
|
85 |
+
|
86 |
+
# Add [SEP] token
|
87 |
+
tokens += [sep_token]
|
88 |
+
token_type_ids = [sequence_a_segment_id] * len(tokens)
|
89 |
+
slot_label_mask += [pad_token_label_id]
|
90 |
+
|
91 |
+
# Add [CLS] token
|
92 |
+
tokens = [cls_token] + tokens
|
93 |
+
token_type_ids = [cls_token_segment_id] + token_type_ids
|
94 |
+
slot_label_mask = [pad_token_label_id] + slot_label_mask
|
95 |
+
|
96 |
+
input_ids = tokenizer.convert_tokens_to_ids(tokens)
|
97 |
+
|
98 |
+
# The mask has 1 for real tokens and 0 for padding tokens. Only real tokens are attended to.
|
99 |
+
attention_mask = [1 if mask_padding_with_zero else 0] * len(input_ids)
|
100 |
+
|
101 |
+
# Zero-pad up to the sequence length.
|
102 |
+
padding_length = args.max_seq_len - len(input_ids)
|
103 |
+
input_ids = input_ids + ([pad_token_id] * padding_length)
|
104 |
+
attention_mask = attention_mask + ([0 if mask_padding_with_zero else 1] * padding_length)
|
105 |
+
token_type_ids = token_type_ids + ([pad_token_segment_id] * padding_length)
|
106 |
+
slot_label_mask = slot_label_mask + ([pad_token_label_id] * padding_length)
|
107 |
+
|
108 |
+
all_input_ids.append(input_ids)
|
109 |
+
all_attention_mask.append(attention_mask)
|
110 |
+
all_token_type_ids.append(token_type_ids)
|
111 |
+
all_slot_label_mask.append(slot_label_mask)
|
112 |
+
|
113 |
+
# Change to Tensor
|
114 |
+
all_input_ids = torch.tensor(all_input_ids, dtype=torch.long)
|
115 |
+
all_attention_mask = torch.tensor(all_attention_mask, dtype=torch.long)
|
116 |
+
all_token_type_ids = torch.tensor(all_token_type_ids, dtype=torch.long)
|
117 |
+
all_slot_label_mask = torch.tensor(all_slot_label_mask, dtype=torch.long)
|
118 |
+
|
119 |
+
dataset = TensorDataset(all_input_ids, all_attention_mask, all_token_type_ids, all_slot_label_mask)
|
120 |
+
|
121 |
+
return dataset
|
122 |
+
|
123 |
+
def predict(text):
|
124 |
+
|
125 |
+
lines = text
|
126 |
+
dataset = convert_input_file_to_tensor_dataset(lines, pred_config, args, tokenizer, pad_token_label_id)
|
127 |
+
|
128 |
+
# Predict
|
129 |
+
sampler = SequentialSampler(dataset)
|
130 |
+
data_loader = DataLoader(dataset, sampler=sampler, batch_size=pred_config.batch_size)
|
131 |
+
|
132 |
+
all_slot_label_mask = None
|
133 |
+
intent_preds = None
|
134 |
+
slot_preds = None
|
135 |
+
|
136 |
+
for batch in tqdm(data_loader, desc="Predicting"):
|
137 |
+
batch = tuple(t.to(device) for t in batch)
|
138 |
+
with torch.no_grad():
|
139 |
+
inputs = {
|
140 |
+
"input_ids": batch[0],
|
141 |
+
"attention_mask": batch[1],
|
142 |
+
"intent_label_ids": None,
|
143 |
+
"slot_labels_ids": None,
|
144 |
+
}
|
145 |
+
if args.model_type != "distilbert":
|
146 |
+
inputs["token_type_ids"] = batch[2]
|
147 |
+
outputs = model(**inputs)
|
148 |
+
_, (intent_logits, slot_logits) = outputs[:2]
|
149 |
+
|
150 |
+
# Intent Prediction
|
151 |
+
if intent_preds is None:
|
152 |
+
intent_preds = intent_logits.detach().cpu().numpy()
|
153 |
+
else:
|
154 |
+
intent_preds = np.append(intent_preds, intent_logits.detach().cpu().numpy(), axis=0)
|
155 |
+
|
156 |
+
# Slot prediction
|
157 |
+
if slot_preds is None:
|
158 |
+
if args.use_crf:
|
159 |
+
# decode() in `torchcrf` returns list with best index directly
|
160 |
+
slot_preds = np.array(model.crf.decode(slot_logits))
|
161 |
+
else:
|
162 |
+
slot_preds = slot_logits.detach().cpu().numpy()
|
163 |
+
all_slot_label_mask = batch[3].detach().cpu().numpy()
|
164 |
+
else:
|
165 |
+
if args.use_crf:
|
166 |
+
slot_preds = np.append(slot_preds, np.array(model.crf.decode(slot_logits)), axis=0)
|
167 |
+
else:
|
168 |
+
slot_preds = np.append(slot_preds, slot_logits.detach().cpu().numpy(), axis=0)
|
169 |
+
all_slot_label_mask = np.append(all_slot_label_mask, batch[3].detach().cpu().numpy(), axis=0)
|
170 |
+
|
171 |
+
intent_preds = np.argmax(intent_preds, axis=1)
|
172 |
+
|
173 |
+
if not args.use_crf:
|
174 |
+
slot_preds = np.argmax(slot_preds, axis=2)
|
175 |
+
|
176 |
+
slot_label_map = {i: label for i, label in enumerate(slot_label_lst)}
|
177 |
+
slot_preds_list = [[] for _ in range(slot_preds.shape[0])]
|
178 |
+
|
179 |
+
for i in range(slot_preds.shape[0]):
|
180 |
+
for j in range(slot_preds.shape[1]):
|
181 |
+
if all_slot_label_mask[i, j] != pad_token_label_id:
|
182 |
+
slot_preds_list[i].append(slot_label_map[slot_preds[i][j]])
|
183 |
+
|
184 |
+
return (lines, slot_preds_list, intent_preds)
|
185 |
+
|
186 |
+
|
187 |
+
def text_analysis(text):
|
188 |
+
text = [text.strip().split()]
|
189 |
+
|
190 |
+
words, slot_preds, intent_pred = predict(text)[0][0], predict(text)[1][0], predict(text)[2][0]
|
191 |
+
|
192 |
+
slot_tokens = []
|
193 |
+
|
194 |
+
for word, pred in zip(words, slot_preds):
|
195 |
+
if pred == 'O':
|
196 |
+
slot_tokens.extend([(word, None), (" ", None)])
|
197 |
+
elif pred[0] == 'I':
|
198 |
+
added_tokens = list(slot_tokens[-2])
|
199 |
+
added_tokens[0] += f' {word}'
|
200 |
+
slot_tokens[-2] = tuple(added_tokens)
|
201 |
+
else:
|
202 |
+
slot_tokens.extend([(word, pred[2:]), (" ", None)])
|
203 |
+
|
204 |
+
intent_label = intent_label_lst[intent_pred]
|
205 |
+
|
206 |
+
return slot_tokens, intent_label
|
207 |
+
|
208 |
+
|
209 |
+
|
210 |
+
if __name__ == "__main__":
|
211 |
+
init_logger()
|
212 |
+
parser = argparse.ArgumentParser()
|
213 |
+
|
214 |
+
# parser.add_argument("--input_file", default="sample_pred_in.txt", type=str, help="Input file for prediction")
|
215 |
+
# parser.add_argument("--output_file", default="sample_pred_out.txt", type=str, help="Output file for prediction")
|
216 |
+
parser.add_argument("--model_dir", default="./atis_model", type=str, help="Path to save, load model")
|
217 |
+
|
218 |
+
parser.add_argument("--batch_size", default=32, type=int, help="Batch size for prediction")
|
219 |
+
parser.add_argument("--no_cuda", action="store_true", help="Avoid using CUDA when available")
|
220 |
+
|
221 |
+
pred_config = parser.parse_args()
|
222 |
+
|
223 |
+
# load model and args
|
224 |
+
args = get_args(pred_config)
|
225 |
+
device = get_device(pred_config)
|
226 |
+
model = load_model(pred_config, args, device)
|
227 |
+
logger.info(args)
|
228 |
+
|
229 |
+
intent_label_lst = get_intent_labels(args)
|
230 |
+
slot_label_lst = get_slot_labels(args)
|
231 |
+
|
232 |
+
# Convert input file to TensorDataset
|
233 |
+
pad_token_label_id = args.ignore_index
|
234 |
+
tokenizer = load_tokenizer(args)
|
235 |
+
|
236 |
+
|
237 |
+
examples = ["tôi muốn bay một chuyến khứ_hồi từ đà_nẵng đến đà_lạt",
|
238 |
+
("giá vé khứ_hồi từ đà_nẵng đến vinh dưới 2 triệu đồng giá vé khứ_hồi từ quy nhơn đến vinh dưới 3 triệu đồng giá vé khứ_hồi từ"
|
239 |
+
" buôn_ma_thuột đến vinh dưới 4 triệu rưỡi"),
|
240 |
+
"cho tôi biết các chuyến bay đến đà_nẵng vào ngày 14 tháng sáu",
|
241 |
+
"những chuyến bay nào khởi_hành từ thành_phố hồ_chí_minh bay đến frankfurt mà nối chuyến ở singapore và hạ_cánh trước 9 giờ tối"]
|
242 |
+
|
243 |
+
demo = gr.Interface(
|
244 |
+
text_analysis,
|
245 |
+
gr.Textbox(placeholder="Enter sentence here...", label="Input"),
|
246 |
+
[gr.HighlightedText(label='Highlighted Output'), gr.Textbox(label='Intent Label')],
|
247 |
+
examples=examples,
|
248 |
+
)
|
249 |
+
|
250 |
+
demo.launch(share=True)
|
main.py
ADDED
@@ -0,0 +1,139 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import argparse
|
2 |
+
|
3 |
+
from data_loader import load_and_cache_examples
|
4 |
+
from trainer import Trainer
|
5 |
+
from utils import MODEL_CLASSES, MODEL_PATH_MAP, init_logger, load_tokenizer, set_seed
|
6 |
+
|
7 |
+
|
8 |
+
def main(args):
|
9 |
+
init_logger()
|
10 |
+
set_seed(args)
|
11 |
+
tokenizer = load_tokenizer(args)
|
12 |
+
|
13 |
+
train_dataset = load_and_cache_examples(args, tokenizer, mode="train")
|
14 |
+
dev_dataset = load_and_cache_examples(args, tokenizer, mode="dev")
|
15 |
+
test_dataset = load_and_cache_examples(args, tokenizer, mode="test")
|
16 |
+
|
17 |
+
trainer = Trainer(args, train_dataset, dev_dataset, test_dataset)
|
18 |
+
|
19 |
+
if args.do_train:
|
20 |
+
trainer.train()
|
21 |
+
|
22 |
+
if args.do_eval:
|
23 |
+
trainer.load_model()
|
24 |
+
trainer.evaluate("test")
|
25 |
+
if args.do_eval_dev:
|
26 |
+
trainer.load_model()
|
27 |
+
trainer.evaluate("dev")
|
28 |
+
|
29 |
+
|
30 |
+
if __name__ == "__main__":
|
31 |
+
parser = argparse.ArgumentParser()
|
32 |
+
|
33 |
+
# parser.add_argument("--task", default=None, required=True, type=str, help="The name of the task to train")
|
34 |
+
parser.add_argument("--model_dir", default=None, required=True, type=str, help="Path to save, load model")
|
35 |
+
parser.add_argument("--data_dir", default="./PhoATIS", type=str, help="The input data dir")
|
36 |
+
parser.add_argument("--intent_label_file", default="intent_label.txt", type=str, help="Intent Label file")
|
37 |
+
parser.add_argument("--slot_label_file", default="slot_label.txt", type=str, help="Slot Label file")
|
38 |
+
|
39 |
+
parser.add_argument(
|
40 |
+
"--model_type",
|
41 |
+
default="phobert",
|
42 |
+
type=str,
|
43 |
+
help="Model type selected in the list: " + ", ".join(MODEL_CLASSES.keys()),
|
44 |
+
)
|
45 |
+
parser.add_argument("--tuning_metric", default="loss", type=str, help="Metrics to tune when training")
|
46 |
+
parser.add_argument("--seed", type=int, default=1, help="random seed for initialization")
|
47 |
+
parser.add_argument("--train_batch_size", default=32, type=int, help="Batch size for training.")
|
48 |
+
parser.add_argument("--eval_batch_size", default=64, type=int, help="Batch size for evaluation.")
|
49 |
+
parser.add_argument(
|
50 |
+
"--max_seq_len", default=50, type=int, help="The maximum total input sequence length after tokenization."
|
51 |
+
)
|
52 |
+
parser.add_argument("--learning_rate", default=5e-5, type=float, help="The initial learning rate for Adam.")
|
53 |
+
parser.add_argument(
|
54 |
+
"--num_train_epochs", default=10.0, type=float, help="Total number of training epochs to perform."
|
55 |
+
)
|
56 |
+
parser.add_argument("--weight_decay", default=0.0, type=float, help="Weight decay if we apply some.")
|
57 |
+
parser.add_argument(
|
58 |
+
"--gradient_accumulation_steps",
|
59 |
+
type=int,
|
60 |
+
default=1,
|
61 |
+
help="Number of updates steps to accumulate before performing a backward/update pass.",
|
62 |
+
)
|
63 |
+
parser.add_argument("--adam_epsilon", default=1e-8, type=float, help="Epsilon for Adam optimizer.")
|
64 |
+
parser.add_argument("--max_grad_norm", default=1.0, type=float, help="Max gradient norm.")
|
65 |
+
parser.add_argument(
|
66 |
+
"--max_steps",
|
67 |
+
default=-1,
|
68 |
+
type=int,
|
69 |
+
help="If > 0: set total number of training steps to perform. Override num_train_epochs.",
|
70 |
+
)
|
71 |
+
parser.add_argument("--warmup_steps", default=0, type=int, help="Linear warmup over warmup_steps.")
|
72 |
+
parser.add_argument("--dropout_rate", default=0.1, type=float, help="Dropout for fully-connected layers")
|
73 |
+
|
74 |
+
parser.add_argument("--logging_steps", type=int, default=200, help="Log every X updates steps.")
|
75 |
+
parser.add_argument("--save_steps", type=int, default=200, help="Save checkpoint every X updates steps.")
|
76 |
+
|
77 |
+
parser.add_argument("--do_train", action="store_true", help="Whether to run training.")
|
78 |
+
parser.add_argument("--do_eval", action="store_true", help="Whether to run eval on the test set.")
|
79 |
+
parser.add_argument("--do_eval_dev", action="store_true", help="Whether to run eval on the dev set.")
|
80 |
+
|
81 |
+
parser.add_argument("--no_cuda", action="store_true", help="Avoid using CUDA when available")
|
82 |
+
|
83 |
+
parser.add_argument(
|
84 |
+
"--ignore_index",
|
85 |
+
default=0,
|
86 |
+
type=int,
|
87 |
+
help="Specifies a target value that is ignored and does not contribute to the input gradient",
|
88 |
+
)
|
89 |
+
|
90 |
+
parser.add_argument("--intent_loss_coef", type=float, default=0.5, help="Coefficient for the intent loss.")
|
91 |
+
parser.add_argument(
|
92 |
+
"--token_level",
|
93 |
+
type=str,
|
94 |
+
default="word-level",
|
95 |
+
help="Tokens are at syllable level or word level (Vietnamese) [word-level, syllable-level]",
|
96 |
+
)
|
97 |
+
parser.add_argument(
|
98 |
+
"--early_stopping",
|
99 |
+
type=int,
|
100 |
+
default=50,
|
101 |
+
help="Number of unincreased validation step to wait for early stopping",
|
102 |
+
)
|
103 |
+
parser.add_argument("--gpu_id", type=int, default=0, help="Select gpu id")
|
104 |
+
# CRF option
|
105 |
+
parser.add_argument("--use_crf", action="store_true", help="Whether to use CRF")
|
106 |
+
# init pretrained
|
107 |
+
parser.add_argument("--pretrained", action="store_true", help="Whether to init model from pretrained base model")
|
108 |
+
parser.add_argument("--pretrained_path", default="./viatis_xlmr_crf", type=str, help="The pretrained model path")
|
109 |
+
|
110 |
+
# Slot-intent interaction
|
111 |
+
parser.add_argument(
|
112 |
+
"--use_intent_context_concat",
|
113 |
+
action="store_true",
|
114 |
+
help="Whether to feed context information of intent into slots vectors (simple concatenation)",
|
115 |
+
)
|
116 |
+
parser.add_argument(
|
117 |
+
"--use_intent_context_attention",
|
118 |
+
action="store_true",
|
119 |
+
help="Whether to feed context information of intent into slots vectors (dot product attention)",
|
120 |
+
)
|
121 |
+
parser.add_argument(
|
122 |
+
"--attention_embedding_size", type=int, default=200, help="hidden size of attention output vector"
|
123 |
+
)
|
124 |
+
|
125 |
+
parser.add_argument(
|
126 |
+
"--slot_pad_label",
|
127 |
+
default="PAD",
|
128 |
+
type=str,
|
129 |
+
help="Pad token for slot label pad (to be ignore when calculate loss)",
|
130 |
+
)
|
131 |
+
parser.add_argument(
|
132 |
+
"--embedding_type", default="soft", type=str, help="Embedding type for intent vector (hard/soft)"
|
133 |
+
)
|
134 |
+
parser.add_argument("--use_attention_mask", action="store_true", help="Whether to use attention mask")
|
135 |
+
|
136 |
+
args = parser.parse_args()
|
137 |
+
|
138 |
+
args.model_name_or_path = MODEL_PATH_MAP[args.model_type]
|
139 |
+
main(args)
|
predict.py
ADDED
@@ -0,0 +1,232 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import argparse
|
2 |
+
import logging
|
3 |
+
import os
|
4 |
+
|
5 |
+
import numpy as np
|
6 |
+
import torch
|
7 |
+
from torch.utils.data import DataLoader, SequentialSampler, TensorDataset
|
8 |
+
from tqdm import tqdm
|
9 |
+
from utils import MODEL_CLASSES, get_intent_labels, get_slot_labels, init_logger, load_tokenizer
|
10 |
+
|
11 |
+
|
12 |
+
logger = logging.getLogger(__name__)
|
13 |
+
|
14 |
+
|
15 |
+
def get_device(pred_config):
|
16 |
+
return "cuda" if torch.cuda.is_available() and not pred_config.no_cuda else "cpu"
|
17 |
+
|
18 |
+
|
19 |
+
def get_args(pred_config):
|
20 |
+
args = torch.load(os.path.join(pred_config.model_dir, "training_args.bin"))
|
21 |
+
|
22 |
+
args.model_dir = 'JointBERT-CRF_PhoBERTencoder'
|
23 |
+
args.data_dir = 'PhoATIS'
|
24 |
+
|
25 |
+
return args
|
26 |
+
|
27 |
+
|
28 |
+
def load_model(pred_config, args, device):
|
29 |
+
# Check whether model exists
|
30 |
+
if not os.path.exists(pred_config.model_dir):
|
31 |
+
raise Exception("Model doesn't exists! Train first!")
|
32 |
+
|
33 |
+
try:
|
34 |
+
model = MODEL_CLASSES[args.model_type][1].from_pretrained(
|
35 |
+
args.model_dir, args=args, intent_label_lst=get_intent_labels(args), slot_label_lst=get_slot_labels(args)
|
36 |
+
)
|
37 |
+
model.to(device)
|
38 |
+
model.eval()
|
39 |
+
logger.info("***** Model Loaded *****")
|
40 |
+
except Exception:
|
41 |
+
raise Exception("Some model files might be missing...")
|
42 |
+
|
43 |
+
return model
|
44 |
+
|
45 |
+
|
46 |
+
def read_input_file(pred_config):
|
47 |
+
lines = []
|
48 |
+
with open(pred_config.input_file, "r", encoding="utf-8") as f:
|
49 |
+
for line in f:
|
50 |
+
line = line.strip()
|
51 |
+
words = line.split()
|
52 |
+
lines.append(words)
|
53 |
+
|
54 |
+
return lines
|
55 |
+
|
56 |
+
|
57 |
+
def convert_input_file_to_tensor_dataset(
|
58 |
+
lines,
|
59 |
+
pred_config,
|
60 |
+
args,
|
61 |
+
tokenizer,
|
62 |
+
pad_token_label_id,
|
63 |
+
cls_token_segment_id=0,
|
64 |
+
pad_token_segment_id=0,
|
65 |
+
sequence_a_segment_id=0,
|
66 |
+
mask_padding_with_zero=True,
|
67 |
+
):
|
68 |
+
# Setting based on the current model type
|
69 |
+
cls_token = tokenizer.cls_token
|
70 |
+
sep_token = tokenizer.sep_token
|
71 |
+
unk_token = tokenizer.unk_token
|
72 |
+
pad_token_id = tokenizer.pad_token_id
|
73 |
+
|
74 |
+
all_input_ids = []
|
75 |
+
all_attention_mask = []
|
76 |
+
all_token_type_ids = []
|
77 |
+
all_slot_label_mask = []
|
78 |
+
|
79 |
+
for words in lines:
|
80 |
+
tokens = []
|
81 |
+
slot_label_mask = []
|
82 |
+
for word in words:
|
83 |
+
word_tokens = tokenizer.tokenize(word)
|
84 |
+
if not word_tokens:
|
85 |
+
word_tokens = [unk_token] # For handling the bad-encoded word
|
86 |
+
tokens.extend(word_tokens)
|
87 |
+
# Use the real label id for the first token of the word, and padding ids for the remaining tokens
|
88 |
+
slot_label_mask.extend([pad_token_label_id + 1] + [pad_token_label_id] * (len(word_tokens) - 1))
|
89 |
+
|
90 |
+
# Account for [CLS] and [SEP]
|
91 |
+
special_tokens_count = 2
|
92 |
+
if len(tokens) > args.max_seq_len - special_tokens_count:
|
93 |
+
tokens = tokens[: (args.max_seq_len - special_tokens_count)]
|
94 |
+
slot_label_mask = slot_label_mask[: (args.max_seq_len - special_tokens_count)]
|
95 |
+
|
96 |
+
# Add [SEP] token
|
97 |
+
tokens += [sep_token]
|
98 |
+
token_type_ids = [sequence_a_segment_id] * len(tokens)
|
99 |
+
slot_label_mask += [pad_token_label_id]
|
100 |
+
|
101 |
+
# Add [CLS] token
|
102 |
+
tokens = [cls_token] + tokens
|
103 |
+
token_type_ids = [cls_token_segment_id] + token_type_ids
|
104 |
+
slot_label_mask = [pad_token_label_id] + slot_label_mask
|
105 |
+
|
106 |
+
input_ids = tokenizer.convert_tokens_to_ids(tokens)
|
107 |
+
|
108 |
+
# The mask has 1 for real tokens and 0 for padding tokens. Only real tokens are attended to.
|
109 |
+
attention_mask = [1 if mask_padding_with_zero else 0] * len(input_ids)
|
110 |
+
|
111 |
+
# Zero-pad up to the sequence length.
|
112 |
+
padding_length = args.max_seq_len - len(input_ids)
|
113 |
+
input_ids = input_ids + ([pad_token_id] * padding_length)
|
114 |
+
attention_mask = attention_mask + ([0 if mask_padding_with_zero else 1] * padding_length)
|
115 |
+
token_type_ids = token_type_ids + ([pad_token_segment_id] * padding_length)
|
116 |
+
slot_label_mask = slot_label_mask + ([pad_token_label_id] * padding_length)
|
117 |
+
|
118 |
+
all_input_ids.append(input_ids)
|
119 |
+
all_attention_mask.append(attention_mask)
|
120 |
+
all_token_type_ids.append(token_type_ids)
|
121 |
+
all_slot_label_mask.append(slot_label_mask)
|
122 |
+
|
123 |
+
# Change to Tensor
|
124 |
+
all_input_ids = torch.tensor(all_input_ids, dtype=torch.long)
|
125 |
+
all_attention_mask = torch.tensor(all_attention_mask, dtype=torch.long)
|
126 |
+
all_token_type_ids = torch.tensor(all_token_type_ids, dtype=torch.long)
|
127 |
+
all_slot_label_mask = torch.tensor(all_slot_label_mask, dtype=torch.long)
|
128 |
+
|
129 |
+
dataset = TensorDataset(all_input_ids, all_attention_mask, all_token_type_ids, all_slot_label_mask)
|
130 |
+
|
131 |
+
return dataset
|
132 |
+
|
133 |
+
|
134 |
+
def predict(pred_config):
|
135 |
+
# load model and args
|
136 |
+
args = get_args(pred_config)
|
137 |
+
device = get_device(pred_config)
|
138 |
+
model = load_model(pred_config, args, device)
|
139 |
+
logger.info(args)
|
140 |
+
|
141 |
+
intent_label_lst = get_intent_labels(args)
|
142 |
+
slot_label_lst = get_slot_labels(args)
|
143 |
+
|
144 |
+
# Convert input file to TensorDataset
|
145 |
+
pad_token_label_id = args.ignore_index
|
146 |
+
tokenizer = load_tokenizer(args)
|
147 |
+
lines = read_input_file(pred_config)
|
148 |
+
dataset = convert_input_file_to_tensor_dataset(lines, pred_config, args, tokenizer, pad_token_label_id)
|
149 |
+
|
150 |
+
# Predict
|
151 |
+
sampler = SequentialSampler(dataset)
|
152 |
+
data_loader = DataLoader(dataset, sampler=sampler, batch_size=pred_config.batch_size)
|
153 |
+
|
154 |
+
all_slot_label_mask = None
|
155 |
+
intent_preds = None
|
156 |
+
slot_preds = None
|
157 |
+
|
158 |
+
for batch in tqdm(data_loader, desc="Predicting"):
|
159 |
+
batch = tuple(t.to(device) for t in batch)
|
160 |
+
with torch.no_grad():
|
161 |
+
inputs = {
|
162 |
+
"input_ids": batch[0],
|
163 |
+
"attention_mask": batch[1],
|
164 |
+
"intent_label_ids": None,
|
165 |
+
"slot_labels_ids": None,
|
166 |
+
}
|
167 |
+
if args.model_type != "distilbert":
|
168 |
+
inputs["token_type_ids"] = batch[2]
|
169 |
+
outputs = model(**inputs)
|
170 |
+
_, (intent_logits, slot_logits) = outputs[:2]
|
171 |
+
|
172 |
+
# Intent Prediction
|
173 |
+
if intent_preds is None:
|
174 |
+
intent_preds = intent_logits.detach().cpu().numpy()
|
175 |
+
else:
|
176 |
+
intent_preds = np.append(intent_preds, intent_logits.detach().cpu().numpy(), axis=0)
|
177 |
+
|
178 |
+
# Slot prediction
|
179 |
+
if slot_preds is None:
|
180 |
+
if args.use_crf:
|
181 |
+
# decode() in `torchcrf` returns list with best index directly
|
182 |
+
slot_preds = np.array(model.crf.decode(slot_logits))
|
183 |
+
else:
|
184 |
+
slot_preds = slot_logits.detach().cpu().numpy()
|
185 |
+
all_slot_label_mask = batch[3].detach().cpu().numpy()
|
186 |
+
else:
|
187 |
+
if args.use_crf:
|
188 |
+
slot_preds = np.append(slot_preds, np.array(model.crf.decode(slot_logits)), axis=0)
|
189 |
+
else:
|
190 |
+
slot_preds = np.append(slot_preds, slot_logits.detach().cpu().numpy(), axis=0)
|
191 |
+
all_slot_label_mask = np.append(all_slot_label_mask, batch[3].detach().cpu().numpy(), axis=0)
|
192 |
+
|
193 |
+
intent_preds = np.argmax(intent_preds, axis=1)
|
194 |
+
|
195 |
+
if not args.use_crf:
|
196 |
+
slot_preds = np.argmax(slot_preds, axis=2)
|
197 |
+
|
198 |
+
slot_label_map = {i: label for i, label in enumerate(slot_label_lst)}
|
199 |
+
slot_preds_list = [[] for _ in range(slot_preds.shape[0])]
|
200 |
+
|
201 |
+
for i in range(slot_preds.shape[0]):
|
202 |
+
for j in range(slot_preds.shape[1]):
|
203 |
+
if all_slot_label_mask[i, j] != pad_token_label_id:
|
204 |
+
slot_preds_list[i].append(slot_label_map[slot_preds[i][j]])
|
205 |
+
|
206 |
+
# Write to output file
|
207 |
+
with open(pred_config.output_file, "w", encoding="utf-8") as f:
|
208 |
+
for words, slot_preds, intent_pred in zip(lines, slot_preds_list, intent_preds):
|
209 |
+
line = ""
|
210 |
+
for word, pred in zip(words, slot_preds):
|
211 |
+
if pred == "O":
|
212 |
+
line = line + word + " "
|
213 |
+
else:
|
214 |
+
line = line + "[{}:{}] ".format(word, pred)
|
215 |
+
f.write("<{}> -> {}\n".format(intent_label_lst[intent_pred], line.strip()))
|
216 |
+
|
217 |
+
logger.info("Prediction Done!")
|
218 |
+
|
219 |
+
|
220 |
+
if __name__ == "__main__":
|
221 |
+
init_logger()
|
222 |
+
parser = argparse.ArgumentParser()
|
223 |
+
|
224 |
+
parser.add_argument("--input_file", default="sample_pred_in.txt", type=str, help="Input file for prediction")
|
225 |
+
parser.add_argument("--output_file", default="sample_pred_out.txt", type=str, help="Output file for prediction")
|
226 |
+
parser.add_argument("--model_dir", default="./atis_model", type=str, help="Path to save, load model")
|
227 |
+
|
228 |
+
parser.add_argument("--batch_size", default=32, type=int, help="Batch size for prediction")
|
229 |
+
parser.add_argument("--no_cuda", action="store_true", help="Avoid using CUDA when available")
|
230 |
+
|
231 |
+
pred_config = parser.parse_args()
|
232 |
+
predict(pred_config)
|
predict.sh
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
python3 predict.py --input_file data/viatis/test/seq.in \
|
2 |
+
--output_file predictions.txt \
|
3 |
+
--model_dir viatis_phobert_crf_attn/4e-5/0.15
|
requirements.txt
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
torch==2.0.0
|
2 |
+
transformers
|
3 |
+
seqeval
|
4 |
+
pytorch-crf
|
5 |
+
tensorflow
|
6 |
+
sentencepiece
|
7 |
+
tensorboard
|
8 |
+
numpy>=1.21.2
|
9 |
+
tqdm
|
10 |
+
typing_extensions
|
11 |
+
protobuf<5,>=3.20.3
|
run_jointBERT-CRF_PhoBERTencoder.sh
ADDED
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
export lr=3e-5
|
2 |
+
export c=0.6
|
3 |
+
export s=100
|
4 |
+
echo "${lr}"
|
5 |
+
export MODEL_DIR=JointBERT-CRF_PhoBERTencoder
|
6 |
+
export MODEL_DIR=$MODEL_DIR"/"$lr"/"$c"/"$s
|
7 |
+
echo "${MODEL_DIR}"
|
8 |
+
python3 main.py --token_level word-level \
|
9 |
+
--model_type phobert \
|
10 |
+
--model_dir $MODEL_DIR \
|
11 |
+
--data_dir PhoATIS \
|
12 |
+
--seed $s \
|
13 |
+
--do_train \
|
14 |
+
--do_eval \
|
15 |
+
--save_steps 140 \
|
16 |
+
--logging_steps 140 \
|
17 |
+
--num_train_epochs 50 \
|
18 |
+
--tuning_metric mean_intent_slot \
|
19 |
+
--use_crf \
|
20 |
+
--gpu_id 0 \
|
21 |
+
--embedding_type soft \
|
22 |
+
--intent_loss_coef $c \
|
23 |
+
--learning_rate $lr
|
run_jointBERT-CRF_XLM-Rencoder.sh
ADDED
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
export lr=4e-5
|
2 |
+
export c=0.45
|
3 |
+
export s=10
|
4 |
+
echo "${lr}"
|
5 |
+
export MODEL_DIR=JointBERT-CRF_XLM-Rencoder
|
6 |
+
export MODEL_DIR=$MODEL_DIR"/"$lr"/"$c"/"$s
|
7 |
+
echo "${MODEL_DIR}"
|
8 |
+
python3 main.py --token_level syllable-level \
|
9 |
+
--model_type xlmr \
|
10 |
+
--model_dir $MODEL_DIR \
|
11 |
+
--data_dir PhoATIS \
|
12 |
+
--seed $s \
|
13 |
+
--do_train \
|
14 |
+
--do_eval \
|
15 |
+
--save_steps 140 \
|
16 |
+
--logging_steps 140 \
|
17 |
+
--num_train_epochs 50 \
|
18 |
+
--tuning_metric mean_intent_slot \
|
19 |
+
--use_crf \
|
20 |
+
--gpu_id 0 \
|
21 |
+
--embedding_type soft \
|
22 |
+
--intent_loss_coef $c \
|
23 |
+
--learning_rate $lr
|
run_jointIDSF_PhoBERTencoder.sh
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#As we initialize JointIDSF from JointBERT, user need to train a base model JointBERT first
|
2 |
+
./run_jointBERT-CRF_PhoBERTencoder.sh
|
3 |
+
#Train JointIDSF
|
4 |
+
export lr=4e-5
|
5 |
+
export c=0.15
|
6 |
+
export s=100
|
7 |
+
echo "${lr}"
|
8 |
+
export MODEL_DIR=JointIDSF_PhoBERTencoder
|
9 |
+
export MODEL_DIR=$MODEL_DIR"/"$lr"/"$c"/"$s
|
10 |
+
echo "${MODEL_DIR}"
|
11 |
+
python3 main.py --token_level word-level \
|
12 |
+
--model_type phobert \
|
13 |
+
--model_dir $MODEL_DIR \
|
14 |
+
--data_dir PhoATIS \
|
15 |
+
--seed $s \
|
16 |
+
--do_train \
|
17 |
+
--do_eval \
|
18 |
+
--save_steps 140 \
|
19 |
+
--logging_steps 140 \
|
20 |
+
--num_train_epochs 50 \
|
21 |
+
--tuning_metric mean_intent_slot \
|
22 |
+
--use_intent_context_attention \
|
23 |
+
--attention_embedding_size 200 \
|
24 |
+
--use_crf \
|
25 |
+
--gpu_id 0 \
|
26 |
+
--embedding_type soft \
|
27 |
+
--intent_loss_coef $c \
|
28 |
+
--pretrained \
|
29 |
+
--pretrained_path JointBERT-CRF_PhoBERTencoder/3e-5/0.6/100 \
|
30 |
+
--learning_rate $lr
|
run_jointIDSF_XLM-Rencoder.sh
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
#As we initialize JointIDSF from JointBERT, user need to train a base model JointBERT first
|
2 |
+
./run_jointBERT-CRF_XLM-Rencoder.sh
|
3 |
+
#Train JointIDSF
|
4 |
+
export lr=3e-5
|
5 |
+
export c=0.25
|
6 |
+
export s=10
|
7 |
+
echo "${lr}"
|
8 |
+
export MODEL_DIR=JointIDSF_XLM-Rencoder
|
9 |
+
export MODEL_DIR=$MODEL_DIR"/"$lr"/"$c"/"$s
|
10 |
+
echo "${MODEL_DIR}"
|
11 |
+
python3 main.py --token_level syllable-level \
|
12 |
+
--model_type xlmr \
|
13 |
+
--model_dir $MODEL_DIR \
|
14 |
+
--data_dir PhoATIS \
|
15 |
+
--seed $s \
|
16 |
+
--do_train \
|
17 |
+
--do_eval \
|
18 |
+
--save_steps 140 \
|
19 |
+
--logging_steps 140 \
|
20 |
+
--num_train_epochs 50 \
|
21 |
+
--tuning_metric mean_intent_slot \
|
22 |
+
--use_intent_context_attention \
|
23 |
+
--attention_embedding_size 200 \
|
24 |
+
--use_crf \
|
25 |
+
--gpu_id 0 \
|
26 |
+
--embedding_type soft \
|
27 |
+
--intent_loss_coef $c \
|
28 |
+
--pretrained \
|
29 |
+
--pretrained_path JointBERT-CRF_XLM-Rencoder/4e-5/0.45/10 \
|
30 |
+
--learning_rate $lr
|
trainer.py
ADDED
@@ -0,0 +1,300 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import logging
|
2 |
+
import os
|
3 |
+
|
4 |
+
import numpy as np
|
5 |
+
import torch
|
6 |
+
from early_stopping import EarlyStopping
|
7 |
+
from torch.utils.data import DataLoader, RandomSampler, SequentialSampler
|
8 |
+
from torch.utils.tensorboard import SummaryWriter
|
9 |
+
from tqdm.auto import tqdm, trange
|
10 |
+
from transformers import AdamW, get_linear_schedule_with_warmup
|
11 |
+
from utils import MODEL_CLASSES, compute_metrics, get_intent_labels, get_slot_labels
|
12 |
+
|
13 |
+
|
14 |
+
logger = logging.getLogger(__name__)
|
15 |
+
|
16 |
+
|
17 |
+
class Trainer(object):
|
18 |
+
def __init__(self, args, train_dataset=None, dev_dataset=None, test_dataset=None):
|
19 |
+
self.args = args
|
20 |
+
self.train_dataset = train_dataset
|
21 |
+
self.dev_dataset = dev_dataset
|
22 |
+
self.test_dataset = test_dataset
|
23 |
+
|
24 |
+
self.intent_label_lst = get_intent_labels(args)
|
25 |
+
self.slot_label_lst = get_slot_labels(args)
|
26 |
+
# Use cross entropy ignore index as padding label id so that only real label ids contribute to the loss later
|
27 |
+
self.pad_token_label_id = args.ignore_index
|
28 |
+
self.config_class, self.model_class, _ = MODEL_CLASSES[args.model_type]
|
29 |
+
# self.config = self.config_class.from_pretrained(model_path, finetuning_task=args.task)
|
30 |
+
|
31 |
+
if args.pretrained:
|
32 |
+
print(args.model_name_or_path)
|
33 |
+
self.model = self.model_class.from_pretrained(
|
34 |
+
args.pretrained_path,
|
35 |
+
args=args,
|
36 |
+
intent_label_lst=self.intent_label_lst,
|
37 |
+
slot_label_lst=self.slot_label_lst,
|
38 |
+
)
|
39 |
+
else:
|
40 |
+
self.config = self.config_class.from_pretrained(args.model_name_or_path, finetuning_task=args.token_level)
|
41 |
+
self.model = self.model_class.from_pretrained(
|
42 |
+
args.model_name_or_path,
|
43 |
+
config=self.config,
|
44 |
+
args=args,
|
45 |
+
intent_label_lst=self.intent_label_lst,
|
46 |
+
slot_label_lst=self.slot_label_lst,
|
47 |
+
)
|
48 |
+
# GPU or CPU
|
49 |
+
torch.cuda.set_device(self.args.gpu_id)
|
50 |
+
print(self.args.gpu_id)
|
51 |
+
print(torch.cuda.current_device())
|
52 |
+
self.device = "cuda" if torch.cuda.is_available() and not args.no_cuda else "cpu"
|
53 |
+
self.model.to(self.device)
|
54 |
+
|
55 |
+
def train(self):
|
56 |
+
train_sampler = RandomSampler(self.train_dataset)
|
57 |
+
train_dataloader = DataLoader(self.train_dataset, sampler=train_sampler, batch_size=self.args.train_batch_size)
|
58 |
+
writer = SummaryWriter(log_dir=self.args.model_dir)
|
59 |
+
if self.args.max_steps > 0:
|
60 |
+
t_total = self.args.max_steps
|
61 |
+
self.args.num_train_epochs = (
|
62 |
+
self.args.max_steps // (len(train_dataloader) // self.args.gradient_accumulation_steps) + 1
|
63 |
+
)
|
64 |
+
else:
|
65 |
+
t_total = len(train_dataloader) // self.args.gradient_accumulation_steps * self.args.num_train_epochs
|
66 |
+
print("check init")
|
67 |
+
results = self.evaluate("dev")
|
68 |
+
print(results)
|
69 |
+
# Prepare optimizer and schedule (linear warmup and decay)
|
70 |
+
no_decay = ["bias", "LayerNorm.weight"]
|
71 |
+
optimizer_grouped_parameters = [
|
72 |
+
{
|
73 |
+
"params": [p for n, p in self.model.named_parameters() if not any(nd in n for nd in no_decay)],
|
74 |
+
"weight_decay": self.args.weight_decay,
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"params": [p for n, p in self.model.named_parameters() if any(nd in n for nd in no_decay)],
|
78 |
+
"weight_decay": 0.0,
|
79 |
+
},
|
80 |
+
]
|
81 |
+
optimizer = AdamW(optimizer_grouped_parameters, lr=self.args.learning_rate, eps=self.args.adam_epsilon)
|
82 |
+
scheduler = get_linear_schedule_with_warmup(
|
83 |
+
optimizer, num_warmup_steps=self.args.warmup_steps, num_training_steps=t_total
|
84 |
+
)
|
85 |
+
|
86 |
+
# Train!
|
87 |
+
logger.info("***** Running training *****")
|
88 |
+
logger.info(" Num examples = %d", len(self.train_dataset))
|
89 |
+
logger.info(" Num Epochs = %d", self.args.num_train_epochs)
|
90 |
+
logger.info(" Total train batch size = %d", self.args.train_batch_size)
|
91 |
+
logger.info(" Gradient Accumulation steps = %d", self.args.gradient_accumulation_steps)
|
92 |
+
logger.info(" Total optimization steps = %d", t_total)
|
93 |
+
logger.info(" Logging steps = %d", self.args.logging_steps)
|
94 |
+
logger.info(" Save steps = %d", self.args.save_steps)
|
95 |
+
|
96 |
+
global_step = 0
|
97 |
+
tr_loss = 0.0
|
98 |
+
self.model.zero_grad()
|
99 |
+
|
100 |
+
train_iterator = trange(int(self.args.num_train_epochs), desc="Epoch")
|
101 |
+
early_stopping = EarlyStopping(patience=self.args.early_stopping, verbose=True)
|
102 |
+
|
103 |
+
for _ in train_iterator:
|
104 |
+
epoch_iterator = tqdm(train_dataloader, desc="Iteration", position=0, leave=True)
|
105 |
+
print("\nEpoch", _)
|
106 |
+
|
107 |
+
for step, batch in enumerate(epoch_iterator):
|
108 |
+
self.model.train()
|
109 |
+
batch = tuple(t.to(self.device) for t in batch) # GPU or CPU
|
110 |
+
|
111 |
+
inputs = {
|
112 |
+
"input_ids": batch[0],
|
113 |
+
"attention_mask": batch[1],
|
114 |
+
"intent_label_ids": batch[3],
|
115 |
+
"slot_labels_ids": batch[4],
|
116 |
+
}
|
117 |
+
if self.args.model_type != "distilbert":
|
118 |
+
inputs["token_type_ids"] = batch[2]
|
119 |
+
outputs = self.model(**inputs)
|
120 |
+
loss = outputs[0]
|
121 |
+
|
122 |
+
if self.args.gradient_accumulation_steps > 1:
|
123 |
+
loss = loss / self.args.gradient_accumulation_steps
|
124 |
+
|
125 |
+
loss.backward()
|
126 |
+
|
127 |
+
tr_loss += loss.item()
|
128 |
+
if (step + 1) % self.args.gradient_accumulation_steps == 0:
|
129 |
+
torch.nn.utils.clip_grad_norm_(self.model.parameters(), self.args.max_grad_norm)
|
130 |
+
|
131 |
+
optimizer.step()
|
132 |
+
scheduler.step() # Update learning rate schedule
|
133 |
+
self.model.zero_grad()
|
134 |
+
global_step += 1
|
135 |
+
|
136 |
+
if self.args.logging_steps > 0 and global_step % self.args.logging_steps == 0:
|
137 |
+
print("\nTuning metrics:", self.args.tuning_metric)
|
138 |
+
results = self.evaluate("dev")
|
139 |
+
writer.add_scalar("Loss/validation", results["loss"], _)
|
140 |
+
writer.add_scalar("Intent Accuracy/validation", results["intent_acc"], _)
|
141 |
+
writer.add_scalar("Slot F1/validation", results["slot_f1"], _)
|
142 |
+
writer.add_scalar("Mean Intent Slot", results["mean_intent_slot"], _)
|
143 |
+
writer.add_scalar("Sentence Accuracy/validation", results["semantic_frame_acc"], _)
|
144 |
+
early_stopping(results[self.args.tuning_metric], self.model, self.args)
|
145 |
+
if early_stopping.early_stop:
|
146 |
+
print("Early stopping")
|
147 |
+
break
|
148 |
+
|
149 |
+
# if self.args.save_steps > 0 and global_step % self.args.save_steps == 0:
|
150 |
+
# self.save_model()
|
151 |
+
|
152 |
+
if 0 < self.args.max_steps < global_step:
|
153 |
+
epoch_iterator.close()
|
154 |
+
break
|
155 |
+
|
156 |
+
if 0 < self.args.max_steps < global_step or early_stopping.early_stop:
|
157 |
+
train_iterator.close()
|
158 |
+
break
|
159 |
+
writer.add_scalar("Loss/train", tr_loss / global_step, _)
|
160 |
+
|
161 |
+
return global_step, tr_loss / global_step
|
162 |
+
|
163 |
+
def write_evaluation_result(self, out_file, results):
|
164 |
+
out_file = self.args.model_dir + "/" + out_file
|
165 |
+
w = open(out_file, "w", encoding="utf-8")
|
166 |
+
w.write("***** Eval results *****\n")
|
167 |
+
for key in sorted(results.keys()):
|
168 |
+
to_write = " {key} = {value}".format(key=key, value=str(results[key]))
|
169 |
+
w.write(to_write)
|
170 |
+
w.write("\n")
|
171 |
+
w.close()
|
172 |
+
|
173 |
+
def evaluate(self, mode):
|
174 |
+
if mode == "test":
|
175 |
+
dataset = self.test_dataset
|
176 |
+
elif mode == "dev":
|
177 |
+
dataset = self.dev_dataset
|
178 |
+
else:
|
179 |
+
raise Exception("Only dev and test dataset available")
|
180 |
+
|
181 |
+
eval_sampler = SequentialSampler(dataset)
|
182 |
+
eval_dataloader = DataLoader(dataset, sampler=eval_sampler, batch_size=self.args.eval_batch_size)
|
183 |
+
|
184 |
+
# Eval!
|
185 |
+
logger.info("***** Running evaluation on %s dataset *****", mode)
|
186 |
+
logger.info(" Num examples = %d", len(dataset))
|
187 |
+
logger.info(" Batch size = %d", self.args.eval_batch_size)
|
188 |
+
eval_loss = 0.0
|
189 |
+
nb_eval_steps = 0
|
190 |
+
intent_preds = None
|
191 |
+
slot_preds = None
|
192 |
+
out_intent_label_ids = None
|
193 |
+
out_slot_labels_ids = None
|
194 |
+
|
195 |
+
self.model.eval()
|
196 |
+
|
197 |
+
for batch in tqdm(eval_dataloader, desc="Evaluating"):
|
198 |
+
batch = tuple(t.to(self.device) for t in batch)
|
199 |
+
with torch.no_grad():
|
200 |
+
inputs = {
|
201 |
+
"input_ids": batch[0],
|
202 |
+
"attention_mask": batch[1],
|
203 |
+
"intent_label_ids": batch[3],
|
204 |
+
"slot_labels_ids": batch[4],
|
205 |
+
}
|
206 |
+
if self.args.model_type != "distilbert":
|
207 |
+
inputs["token_type_ids"] = batch[2]
|
208 |
+
outputs = self.model(**inputs)
|
209 |
+
tmp_eval_loss, (intent_logits, slot_logits) = outputs[:2]
|
210 |
+
|
211 |
+
eval_loss += tmp_eval_loss.mean().item()
|
212 |
+
nb_eval_steps += 1
|
213 |
+
|
214 |
+
# Intent prediction
|
215 |
+
if intent_preds is None:
|
216 |
+
intent_preds = intent_logits.detach().cpu().numpy()
|
217 |
+
out_intent_label_ids = inputs["intent_label_ids"].detach().cpu().numpy()
|
218 |
+
else:
|
219 |
+
intent_preds = np.append(intent_preds, intent_logits.detach().cpu().numpy(), axis=0)
|
220 |
+
out_intent_label_ids = np.append(
|
221 |
+
out_intent_label_ids, inputs["intent_label_ids"].detach().cpu().numpy(), axis=0
|
222 |
+
)
|
223 |
+
|
224 |
+
# Slot prediction
|
225 |
+
if slot_preds is None:
|
226 |
+
if self.args.use_crf:
|
227 |
+
# decode() in `torchcrf` returns list with best index directly
|
228 |
+
slot_preds = np.array(self.model.crf.decode(slot_logits))
|
229 |
+
else:
|
230 |
+
slot_preds = slot_logits.detach().cpu().numpy()
|
231 |
+
|
232 |
+
out_slot_labels_ids = inputs["slot_labels_ids"].detach().cpu().numpy()
|
233 |
+
else:
|
234 |
+
if self.args.use_crf:
|
235 |
+
slot_preds = np.append(slot_preds, np.array(self.model.crf.decode(slot_logits)), axis=0)
|
236 |
+
else:
|
237 |
+
slot_preds = np.append(slot_preds, slot_logits.detach().cpu().numpy(), axis=0)
|
238 |
+
|
239 |
+
out_slot_labels_ids = np.append(
|
240 |
+
out_slot_labels_ids, inputs["slot_labels_ids"].detach().cpu().numpy(), axis=0
|
241 |
+
)
|
242 |
+
|
243 |
+
eval_loss = eval_loss / nb_eval_steps
|
244 |
+
results = {"loss": eval_loss}
|
245 |
+
|
246 |
+
# Intent result
|
247 |
+
intent_preds = np.argmax(intent_preds, axis=1)
|
248 |
+
|
249 |
+
# Slot result
|
250 |
+
if not self.args.use_crf:
|
251 |
+
slot_preds = np.argmax(slot_preds, axis=2)
|
252 |
+
slot_label_map = {i: label for i, label in enumerate(self.slot_label_lst)}
|
253 |
+
out_slot_label_list = [[] for _ in range(out_slot_labels_ids.shape[0])]
|
254 |
+
slot_preds_list = [[] for _ in range(out_slot_labels_ids.shape[0])]
|
255 |
+
|
256 |
+
for i in range(out_slot_labels_ids.shape[0]):
|
257 |
+
for j in range(out_slot_labels_ids.shape[1]):
|
258 |
+
if out_slot_labels_ids[i, j] != self.pad_token_label_id:
|
259 |
+
out_slot_label_list[i].append(slot_label_map[out_slot_labels_ids[i][j]])
|
260 |
+
slot_preds_list[i].append(slot_label_map[slot_preds[i][j]])
|
261 |
+
|
262 |
+
total_result = compute_metrics(intent_preds, out_intent_label_ids, slot_preds_list, out_slot_label_list)
|
263 |
+
results.update(total_result)
|
264 |
+
|
265 |
+
logger.info("***** Eval results *****")
|
266 |
+
for key in sorted(results.keys()):
|
267 |
+
logger.info(" %s = %s", key, str(results[key]))
|
268 |
+
if mode == "test":
|
269 |
+
self.write_evaluation_result("eval_test_results.txt", results)
|
270 |
+
elif mode == "dev":
|
271 |
+
self.write_evaluation_result("eval_dev_results.txt", results)
|
272 |
+
return results
|
273 |
+
|
274 |
+
def save_model(self):
|
275 |
+
# Save model checkpoint (Overwrite)
|
276 |
+
if not os.path.exists(self.args.model_dir):
|
277 |
+
os.makedirs(self.args.model_dir)
|
278 |
+
model_to_save = self.model.module if hasattr(self.model, "module") else self.model
|
279 |
+
model_to_save.save_pretrained(self.args.model_dir)
|
280 |
+
|
281 |
+
# Save training arguments together with the trained model
|
282 |
+
torch.save(self.args, os.path.join(self.args.model_dir, "training_args.bin"))
|
283 |
+
logger.info("Saving model checkpoint to %s", self.args.model_dir)
|
284 |
+
|
285 |
+
def load_model(self):
|
286 |
+
# Check whether model exists
|
287 |
+
if not os.path.exists(self.args.model_dir):
|
288 |
+
raise Exception("Model doesn't exists! Train first!")
|
289 |
+
|
290 |
+
try:
|
291 |
+
self.model = self.model_class.from_pretrained(
|
292 |
+
self.args.model_dir,
|
293 |
+
args=self.args,
|
294 |
+
intent_label_lst=self.intent_label_lst,
|
295 |
+
slot_label_lst=self.slot_label_lst,
|
296 |
+
)
|
297 |
+
self.model.to(self.device)
|
298 |
+
logger.info("***** Model Loaded *****")
|
299 |
+
except Exception:
|
300 |
+
raise Exception("Some model files might be missing...")
|
utils.py
ADDED
@@ -0,0 +1,115 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import logging
|
2 |
+
import os
|
3 |
+
import random
|
4 |
+
|
5 |
+
import numpy as np
|
6 |
+
import torch
|
7 |
+
from model import JointPhoBERT, JointXLMR
|
8 |
+
from seqeval.metrics import f1_score, precision_score, recall_score
|
9 |
+
from transformers import (
|
10 |
+
AutoTokenizer,
|
11 |
+
RobertaConfig,
|
12 |
+
XLMRobertaConfig,
|
13 |
+
XLMRobertaTokenizer,
|
14 |
+
)
|
15 |
+
|
16 |
+
|
17 |
+
MODEL_CLASSES = {
|
18 |
+
"xlmr": (XLMRobertaConfig, JointXLMR, XLMRobertaTokenizer),
|
19 |
+
"phobert": (RobertaConfig, JointPhoBERT, AutoTokenizer),
|
20 |
+
}
|
21 |
+
|
22 |
+
MODEL_PATH_MAP = {
|
23 |
+
"xlmr": "xlm-roberta-base",
|
24 |
+
"phobert": "vinai/phobert-base",
|
25 |
+
}
|
26 |
+
|
27 |
+
|
28 |
+
def get_intent_labels(args):
|
29 |
+
return [
|
30 |
+
label.strip()
|
31 |
+
for label in open(os.path.join(args.data_dir, args.token_level, args.intent_label_file), "r", encoding="utf-8")
|
32 |
+
]
|
33 |
+
|
34 |
+
|
35 |
+
def get_slot_labels(args):
|
36 |
+
return [
|
37 |
+
label.strip()
|
38 |
+
for label in open(os.path.join(args.data_dir, args.token_level, args.slot_label_file), "r", encoding="utf-8")
|
39 |
+
]
|
40 |
+
|
41 |
+
|
42 |
+
def load_tokenizer(args):
|
43 |
+
return MODEL_CLASSES[args.model_type][2].from_pretrained(args.model_name_or_path)
|
44 |
+
|
45 |
+
|
46 |
+
def init_logger():
|
47 |
+
logging.basicConfig(
|
48 |
+
format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
|
49 |
+
datefmt="%m/%d/%Y %H:%M:%S",
|
50 |
+
level=logging.INFO,
|
51 |
+
)
|
52 |
+
|
53 |
+
|
54 |
+
def set_seed(args):
|
55 |
+
random.seed(args.seed)
|
56 |
+
np.random.seed(args.seed)
|
57 |
+
torch.manual_seed(args.seed)
|
58 |
+
if not args.no_cuda and torch.cuda.is_available():
|
59 |
+
torch.cuda.manual_seed_all(args.seed)
|
60 |
+
|
61 |
+
|
62 |
+
def compute_metrics(intent_preds, intent_labels, slot_preds, slot_labels):
|
63 |
+
assert len(intent_preds) == len(intent_labels) == len(slot_preds) == len(slot_labels)
|
64 |
+
results = {}
|
65 |
+
intent_result = get_intent_acc(intent_preds, intent_labels)
|
66 |
+
slot_result = get_slot_metrics(slot_preds, slot_labels)
|
67 |
+
sementic_result = get_sentence_frame_acc(intent_preds, intent_labels, slot_preds, slot_labels)
|
68 |
+
|
69 |
+
mean_intent_slot = (intent_result["intent_acc"] + slot_result["slot_f1"]) / 2
|
70 |
+
|
71 |
+
results.update(intent_result)
|
72 |
+
results.update(slot_result)
|
73 |
+
results.update(sementic_result)
|
74 |
+
results["mean_intent_slot"] = mean_intent_slot
|
75 |
+
|
76 |
+
return results
|
77 |
+
|
78 |
+
|
79 |
+
def get_slot_metrics(preds, labels):
|
80 |
+
assert len(preds) == len(labels)
|
81 |
+
return {
|
82 |
+
"slot_precision": precision_score(labels, preds),
|
83 |
+
"slot_recall": recall_score(labels, preds),
|
84 |
+
"slot_f1": f1_score(labels, preds),
|
85 |
+
}
|
86 |
+
|
87 |
+
|
88 |
+
def get_intent_acc(preds, labels):
|
89 |
+
acc = (preds == labels).mean()
|
90 |
+
return {"intent_acc": acc}
|
91 |
+
|
92 |
+
|
93 |
+
def read_prediction_text(args):
|
94 |
+
return [text.strip() for text in open(os.path.join(args.pred_dir, args.pred_input_file), "r", encoding="utf-8")]
|
95 |
+
|
96 |
+
|
97 |
+
def get_sentence_frame_acc(intent_preds, intent_labels, slot_preds, slot_labels):
|
98 |
+
"""For the cases that intent and all the slots are correct (in one sentence)"""
|
99 |
+
# Get the intent comparison result
|
100 |
+
intent_result = intent_preds == intent_labels
|
101 |
+
|
102 |
+
# Get the slot comparision result
|
103 |
+
slot_result = []
|
104 |
+
for preds, labels in zip(slot_preds, slot_labels):
|
105 |
+
assert len(preds) == len(labels)
|
106 |
+
one_sent_result = True
|
107 |
+
for p, l in zip(preds, labels):
|
108 |
+
if p != l:
|
109 |
+
one_sent_result = False
|
110 |
+
break
|
111 |
+
slot_result.append(one_sent_result)
|
112 |
+
slot_result = np.array(slot_result)
|
113 |
+
|
114 |
+
semantic_acc = np.multiply(intent_result, slot_result).mean()
|
115 |
+
return {"semantic_frame_acc": semantic_acc}
|