letfd commited on
Commit
992f9d4
·
1 Parent(s): d5b6025

Upload BertForTokenClassification

Browse files
Files changed (2) hide show
  1. config.json +315 -0
  2. pytorch_model.bin +3 -0
config.json ADDED
@@ -0,0 +1,315 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "KoichiYasuoka/bert-base-russian-upos",
3
+ "architectures": [
4
+ "BertForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "directionality": "bidi",
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "id2label": {
13
+ "0": "LABEL_0",
14
+ "1": "LABEL_1",
15
+ "2": "LABEL_2",
16
+ "3": "LABEL_3",
17
+ "4": "LABEL_4",
18
+ "5": "LABEL_5"
19
+ },
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 3072,
22
+ "label2id": {
23
+ "LABEL_0": 0,
24
+ "LABEL_1": 1,
25
+ "LABEL_2": 2,
26
+ "LABEL_3": 3,
27
+ "LABEL_4": 4,
28
+ "LABEL_5": 5
29
+ },
30
+ "layer_norm_eps": 1e-12,
31
+ "max_position_embeddings": 512,
32
+ "model_type": "bert",
33
+ "num_attention_heads": 12,
34
+ "num_hidden_layers": 12,
35
+ "output_past": true,
36
+ "pad_token_id": 0,
37
+ "pooler_fc_size": 768,
38
+ "pooler_num_attention_heads": 12,
39
+ "pooler_num_fc_layers": 3,
40
+ "pooler_size_per_head": 128,
41
+ "pooler_type": "first_token_transform",
42
+ "position_embedding_type": "absolute",
43
+ "task_specific_params": {
44
+ "upos_multiword": {
45
+ "ADJ+ADJ": {
46
+ "\u0432\u043e\u0435\u043d\u043d\u043e\u043c\u043e\u0440\u0441\u043a\u0430\u044f": [
47
+ "\u0432\u043e\u0435\u043d\u043d\u043e",
48
+ "\u043c\u043e\u0440\u0441\u043a\u0430\u044f"
49
+ ],
50
+ "\u0432\u043e\u0435\u043d\u043d\u043e\u043c\u043e\u0440\u0441\u043a\u0443\u044e": [
51
+ "\u0432\u043e\u0435\u043d\u043d\u043e",
52
+ "\u043c\u043e\u0440\u0441\u043a\u0443\u044e"
53
+ ],
54
+ "\u0433\u043e\u0440\u043d\u043e\u043c\u0435\u0442\u0430\u043b\u043b\u0443\u0440\u0433\u0438\u0447\u0435\u0441\u043a\u0438\u0439": [
55
+ "\u0433\u043e\u0440\u043d\u043e",
56
+ "\u043c\u0435\u0442\u0430\u043b\u043b\u0443\u0440\u0433\u0438\u0447\u0435\u0441\u043a\u0438\u0439"
57
+ ],
58
+ "\u0434\u043e\u043b\u0433\u043e\u0441\u0440\u043e\u0447\u043d\u044b\u043c": [
59
+ "\u0434\u043e\u043b\u0433\u043e",
60
+ "\u0441\u0440\u043e\u0447\u043d\u044b\u043c"
61
+ ],
62
+ "\u043a\u043e\u0441\u043c\u043e\u0437\u0435\u043c\u043d\u044b\u0445": [
63
+ "\u043a\u043e\u0441\u043c\u043e",
64
+ "\u0437\u0435\u043c\u043d\u044b\u0445"
65
+ ],
66
+ "\u0441\u0443\u0434\u0435\u0431\u043d\u043e\u043c\u0435\u0434\u0438\u0446\u0438\u043d\u0441\u043a\u043e\u0439": [
67
+ "\u0441\u0443\u0434\u0435\u0431\u043d\u043e",
68
+ "\u043c\u0435\u0434\u0438\u0446\u0438\u043d\u0441\u043a\u043e\u0439"
69
+ ]
70
+ },
71
+ "ADJ+NOUN": {
72
+ "\u0420\u043e\u0441\u0433\u043e\u0441\u0441\u0442\u0440\u0430\u0445\u0430": [
73
+ "\u0420\u043e\u0441",
74
+ "\u0433\u043e\u0441\u0441\u0442\u0440\u0430\u0445\u0430"
75
+ ],
76
+ "\u0420\u043e\u0441\u043a\u0443\u043b\u044c\u0442\u0443\u0440\u044b": [
77
+ "\u0420\u043e\u0441",
78
+ "\u043a\u0443\u043b\u044c\u0442\u0443\u0440\u044b"
79
+ ],
80
+ "\u0433\u043e\u0441\u0437\u0430\u043a\u0443\u043f\u043e\u043a": [
81
+ "\u0433\u043e\u0441",
82
+ "\u0437\u0430\u043a\u0443\u043f\u043e\u043a"
83
+ ],
84
+ "\u0433\u043e\u0441\u043a\u043e\u0440\u043f\u043e\u0440\u0430\u0446\u0438\u044f\u043c": [
85
+ "\u0433\u043e\u0441",
86
+ "\u043a\u043e\u0440\u043f\u043e\u0440\u0430\u0446\u0438\u044f\u043c"
87
+ ],
88
+ "\u0433\u043e\u0441\u043f\u043e\u0434\u0434\u0435\u0440\u0436\u043a\u0430": [
89
+ "\u0433\u043e\u0441",
90
+ "\u043f\u043e\u0434\u0434\u0435\u0440\u0436\u043a\u0430"
91
+ ],
92
+ "\u0433\u043e\u0441\u0441\u043b\u0443\u0436\u0431\u044b": [
93
+ "\u0433\u043e\u0441",
94
+ "\u0441\u043b\u0443\u0436\u0431\u044b"
95
+ ],
96
+ "\u043c\u0435\u0434\u0443\u0447\u0440\u0435\u0436\u0434\u0435\u043d\u0438\u0438": [
97
+ "\u043c\u0435\u0434",
98
+ "\u0443\u0447\u0440\u0435\u0436\u0434\u0435\u043d\u0438\u0438"
99
+ ],
100
+ "\u0441\u0435\u043b\u044c\u0445\u043e\u0437\u043f\u0440\u043e\u0434\u0443\u043a\u0446\u0438\u0438": [
101
+ "\u0441\u0435\u043b\u044c\u0445\u043e\u0437",
102
+ "\u043f\u0440\u043e\u0434\u0443\u043a\u0446\u0438\u0438"
103
+ ]
104
+ },
105
+ "ADJ+VERB": {
106
+ "\u041b\u0435\u0432\u043e\u043d\u0430\u0441\u0442\u0440\u043e\u0435\u043d\u043d\u044b\u0435": [
107
+ "\u041b\u0435\u0432\u043e",
108
+ "\u043d\u0430\u0441\u0442\u0440\u043e\u0435\u043d\u043d\u044b\u0435"
109
+ ]
110
+ },
111
+ "ADP+DET": {
112
+ "\u0441\u0434\u0440\u0443\u0433\u0438\u043c\u0438": [
113
+ "\u0441",
114
+ "\u0434\u0440\u0443\u0433\u0438\u043c\u0438"
115
+ ],
116
+ "\u0441\u0442\u0435\u0445": [
117
+ "\u0441",
118
+ "\u0442\u0435\u0445"
119
+ ]
120
+ },
121
+ "ADP+PRON": {
122
+ "\u0441\u043d\u0438\u043c": [
123
+ "\u0441",
124
+ "\u043d\u0438\u043c"
125
+ ]
126
+ },
127
+ "ADV+VERB": {
128
+ "\u043d\u0435\u0433\u0434\u0435\u0442\u0440\u0435\u043d\u0438\u0440\u043e\u0432\u0430\u0442\u044c": [
129
+ "\u043d\u0435\u0433\u0434\u0435",
130
+ "\u0442\u0440\u0435\u043d\u0438\u0440\u043e\u0432\u0430\u0442\u044c"
131
+ ]
132
+ },
133
+ "NUM+NOUN": {
134
+ "\u041f\u043e\u043b\u043b\u0438\u0442\u0440\u0430": [
135
+ "\u041f\u043e\u043b",
136
+ "\u043b\u0438\u0442\u0440\u0430"
137
+ ],
138
+ "\u043f\u043e\u043b\u0432\u0435\u043a\u0430": [
139
+ "\u043f\u043e\u043b",
140
+ "\u0432\u0435\u043a\u0430"
141
+ ],
142
+ "\u043f\u043e\u043b\u0433\u043e\u0434\u0430": [
143
+ "\u043f\u043e\u043b",
144
+ "\u0433\u043e\u0434\u0430"
145
+ ],
146
+ "\u043f\u043e\u043b\u0434\u043d\u044f": [
147
+ "\u043f\u043e\u043b",
148
+ "\u0434\u043d\u044f"
149
+ ],
150
+ "\u043f\u043e\u043b\u0434\u044e\u0436\u0438\u043d\u044b": [
151
+ "\u043f\u043e\u043b",
152
+ "\u0434\u044e\u0436\u0438\u043d\u044b"
153
+ ],
154
+ "\u043f\u043e\u043b\u043a\u0438\u043b\u043e": [
155
+ "\u043f\u043e\u043b",
156
+ "\u043a\u0438\u043b\u043e"
157
+ ],
158
+ "\u043f\u043e\u043b\u043a\u0438\u043b\u043e\u0433\u0440\u0430\u043c\u043c\u0430": [
159
+ "\u043f\u043e\u043b",
160
+ "\u043a\u0438\u043b\u043e\u0433\u0440\u0430\u043c\u043c\u0430"
161
+ ],
162
+ "\u043f\u043e\u043b\u043a\u043e\u0442\u043b\u0435\u0442\u044b": [
163
+ "\u043f\u043e\u043b",
164
+ "\u043a\u043e\u0442\u043b\u0435\u0442\u044b"
165
+ ],
166
+ "\u043f\u043e\u043b\u043a\u0443\u0441\u043e\u0447\u043a\u0430": [
167
+ "\u043f\u043e\u043b",
168
+ "\u043a\u0443\u0441\u043e\u0447\u043a\u0430"
169
+ ],
170
+ "\u043f\u043e\u043b\u043c\u0435\u0441\u044f\u0446\u0430": [
171
+ "\u043f\u043e\u043b",
172
+ "\u043c\u0435\u0441\u044f\u0446\u0430"
173
+ ],
174
+ "\u043f\u043e\u043b\u043c\u0435\u0442\u0440\u0430": [
175
+ "\u043f\u043e\u043b",
176
+ "\u043c\u0435\u0442\u0440\u0430"
177
+ ],
178
+ "\u043f\u043e\u043b\u043c\u0435\u0448\u043a\u0430": [
179
+ "\u043f\u043e\u043b",
180
+ "\u043c\u0435\u0448\u043a\u0430"
181
+ ],
182
+ "\u043f\u043e\u043b\u043c\u0438\u043b\u043b\u0438\u0430\u0440\u0434\u0430": [
183
+ "\u043f\u043e\u043b",
184
+ "\u043c\u0438\u043b\u043b\u0438\u0430\u0440\u0434\u0430"
185
+ ],
186
+ "\u043f\u043e\u043b\u043c\u0438\u043b\u043b\u0438\u043e\u043d\u0430": [
187
+ "\u043f\u043e\u043b",
188
+ "\u043c\u0438\u043b\u043b\u0438\u043e\u043d\u0430"
189
+ ],
190
+ "\u043f\u043e\u043b\u043c\u0438\u043d\u0443\u0442\u044b": [
191
+ "\u043f\u043e\u043b",
192
+ "\u043c\u0438\u043d\u0443\u0442\u044b"
193
+ ],
194
+ "\u043f\u043e\u043b\u043c\u044b\u0441\u043b\u0438": [
195
+ "\u043f\u043e\u043b",
196
+ "\u043c\u044b\u0441\u043b\u0438"
197
+ ],
198
+ "\u043f\u043e\u043b\u043f\u0443\u0442\u0438": [
199
+ "\u043f\u043e\u043b",
200
+ "\u043f\u0443\u0442\u0438"
201
+ ],
202
+ "\u043f\u043e\u043b\u0447\u0430\u0441\u0430": [
203
+ "\u043f\u043e\u043b",
204
+ "\u0447\u0430\u0441\u0430"
205
+ ]
206
+ },
207
+ "NUM+SYM": {
208
+ "20\u00b0": [
209
+ "20",
210
+ "\u00b0"
211
+ ],
212
+ "2\u00b0": [
213
+ "2",
214
+ "\u00b0"
215
+ ],
216
+ "90\u00b0": [
217
+ "90",
218
+ "\u00b0"
219
+ ]
220
+ },
221
+ "PART+PRON": {
222
+ "\u0434\u0430\u0442\u043e": [
223
+ "\u0434\u0430",
224
+ "\u0442\u043e"
225
+ ]
226
+ },
227
+ "SYM+NUM": {
228
+ "\u21161": [
229
+ "\u2116",
230
+ "1"
231
+ ],
232
+ "\u2116159": [
233
+ "\u2116",
234
+ "159"
235
+ ],
236
+ "\u211619": [
237
+ "\u2116",
238
+ "19"
239
+ ],
240
+ "\u21162": [
241
+ "\u2116",
242
+ "2"
243
+ ],
244
+ "\u211636": [
245
+ "\u2116",
246
+ "36"
247
+ ],
248
+ "\u21165": [
249
+ "\u2116",
250
+ "5"
251
+ ]
252
+ },
253
+ "VERB+ADV": {
254
+ "\u041d\u0435\u043a\u043e\u0433\u0434\u0430": [
255
+ "\u041d\u0435",
256
+ "\u043a\u043e\u0433\u0434\u0430"
257
+ ],
258
+ "\u043d\u0435\u0437\u0430\u0447\u0435\u043c": [
259
+ "\u043d\u0435",
260
+ "\u0437\u0430\u0447\u0435\u043c"
261
+ ],
262
+ "\u043d\u0435\u043a\u043e\u0433\u0434\u0430": [
263
+ "\u043d\u0435",
264
+ "\u043a\u043e\u0433\u0434\u0430"
265
+ ],
266
+ "\u043d\u0435\u043a\u0443\u0434\u0430": [
267
+ "\u043d\u0435",
268
+ "\u043a\u0443\u0434\u0430"
269
+ ]
270
+ },
271
+ "VERB+PRON": {
272
+ "\u041d\u0435\u043a\u043e\u043c\u0443": [
273
+ "\u041d\u0435",
274
+ "\u043a\u043e\u043c\u0443"
275
+ ],
276
+ "\u043d\u0435\u043a\u043e\u0433\u043e": [
277
+ "\u043d\u0435",
278
+ "\u043a\u043e\u0433\u043e"
279
+ ],
280
+ "\u043d\u0435\u043a\u043e\u043c\u0443": [
281
+ "\u043d\u0435",
282
+ "\u043a\u043e\u043c\u0443"
283
+ ],
284
+ "\u043d\u0435\u0447\u0435\u0433\u043e": [
285
+ "\u043d\u0435",
286
+ "\u0447\u0435\u0433\u043e"
287
+ ],
288
+ "\u043d\u0435\u0447\u0435\u043c": [
289
+ "\u043d\u0435",
290
+ "\u0447\u0435\u043c"
291
+ ],
292
+ "\u043d\u0435\u0447\u0435\u043c\u0443": [
293
+ "\u043d\u0435",
294
+ "\u0447\u0435\u043c\u0443"
295
+ ]
296
+ },
297
+ "X+X": {
298
+ "xxx": [
299
+ "x",
300
+ "xx"
301
+ ],
302
+ "xxxx": [
303
+ "x",
304
+ "xxx"
305
+ ]
306
+ }
307
+ }
308
+ },
309
+ "tokenizer_class": "BertTokenizerFast",
310
+ "torch_dtype": "float32",
311
+ "transformers_version": "4.24.0",
312
+ "type_vocab_size": 2,
313
+ "use_cache": true,
314
+ "vocab_size": 119547
315
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6cb309bae0697a0b2b056171cb4b4c4704acdfe86258dd1a4bf252d25e53b41
3
+ size 709141997