crodri commited on
Commit
d5668b6
1 Parent(s): 2f79a7e

initial upload for better distribution and instalation

Browse files
.gitattributes CHANGED
@@ -32,3 +32,10 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ ca_anonimization_core_lg-any-py3-none-any.whl filter=lfs diff=lfs merge=lfs -text
36
+ parser/model filter=lfs diff=lfs merge=lfs -text
37
+ vocab/vectors filter=lfs diff=lfs merge=lfs -text
38
+ vocab/key2row filter=lfs diff=lfs merge=lfs -text
39
+ ner/model filter=lfs diff=lfs merge=lfs -text
40
+ tok2vec/model filter=lfs diff=lfs merge=lfs -text
41
+ morphologizer/model filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - spacy
4
+ - token-classification
5
+ language:
6
+ - ca
7
+ license: mit
8
+ model-index:
9
+ - name: ca_anonimization_core_lg
10
+ results:
11
+ - task:
12
+ name: NER
13
+ type: token-classification
14
+ metrics:
15
+ - name: NER Precision
16
+ type: precision
17
+ value: 0.746031746
18
+ - name: NER Recall
19
+ type: recall
20
+ value: 0.6438356164
21
+ - name: NER F Score
22
+ type: f_score
23
+ value: 0.6911764706
24
+ - task:
25
+ name: POS
26
+ type: token-classification
27
+ metrics:
28
+ - name: POS (UPOS) Accuracy
29
+ type: accuracy
30
+ value: 0.0
31
+ - task:
32
+ name: MORPH
33
+ type: token-classification
34
+ metrics:
35
+ - name: Morph (UFeats) Accuracy
36
+ type: accuracy
37
+ value: 0.0
38
+ - task:
39
+ name: LEMMA
40
+ type: token-classification
41
+ metrics:
42
+ - name: Lemma Accuracy
43
+ type: accuracy
44
+ value: 0.0
45
+ - task:
46
+ name: UNLABELED_DEPENDENCIES
47
+ type: token-classification
48
+ metrics:
49
+ - name: Unlabeled Attachment Score (UAS)
50
+ type: f_score
51
+ value: 0.0
52
+ - task:
53
+ name: LABELED_DEPENDENCIES
54
+ type: token-classification
55
+ metrics:
56
+ - name: Labeled Attachment Score (LAS)
57
+ type: f_score
58
+ value: 0.0
59
+ - task:
60
+ name: SENTS
61
+ type: token-classification
62
+ metrics:
63
+ - name: Sentences F-Score
64
+ type: f_score
65
+ value: 0.0
66
+ ---
attribute_ruler/patterns ADDED
Binary file (972 Bytes). View file
 
ca_anonimization_core_lg-any-py3-none-any.whl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a2879b91453a33ae00f644f4532098de1cf6978c242f8bc1881515ed4089422
3
+ size 587786958
config.cfg ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [paths]
2
+ train = null
3
+ dev = null
4
+ vectors = null
5
+ init_tok2vec = null
6
+
7
+ [system]
8
+ gpu_allocator = null
9
+ seed = 0
10
+
11
+ [nlp]
12
+ lang = "ca"
13
+ pipeline = ["tok2vec","morphologizer","parser","attribute_ruler","lemmatizer","ner"]
14
+ disabled = []
15
+ before_creation = null
16
+ after_creation = null
17
+ after_pipeline_creation = null
18
+ batch_size = 256
19
+ tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
20
+
21
+ [components]
22
+
23
+ [components.attribute_ruler]
24
+ factory = "attribute_ruler"
25
+ scorer = {"@scorers":"spacy.attribute_ruler_scorer.v1"}
26
+ validate = false
27
+
28
+ [components.lemmatizer]
29
+ factory = "lemmatizer"
30
+ mode = "rule"
31
+ model = null
32
+ overwrite = false
33
+ scorer = {"@scorers":"spacy.lemmatizer_scorer.v1"}
34
+
35
+ [components.morphologizer]
36
+ factory = "morphologizer"
37
+ extend = false
38
+ overwrite = true
39
+ scorer = {"@scorers":"spacy.morphologizer_scorer.v1"}
40
+
41
+ [components.morphologizer.model]
42
+ @architectures = "spacy.Tagger.v1"
43
+ nO = null
44
+
45
+ [components.morphologizer.model.tok2vec]
46
+ @architectures = "spacy.Tok2Vec.v2"
47
+
48
+ [components.morphologizer.model.tok2vec.embed]
49
+ @architectures = "spacy.MultiHashEmbed.v2"
50
+ width = 96
51
+ attrs = ["NORM","PREFIX","SUFFIX","SHAPE","SPACY"]
52
+ rows = [5000,2500,2500,2500,100]
53
+ include_static_vectors = true
54
+
55
+ [components.morphologizer.model.tok2vec.encode]
56
+ @architectures = "spacy.MaxoutWindowEncoder.v2"
57
+ width = 96
58
+ depth = 4
59
+ window_size = 1
60
+ maxout_pieces = 3
61
+
62
+ [components.ner]
63
+ factory = "ner"
64
+ incorrect_spans_key = null
65
+ moves = null
66
+ scorer = {"@scorers":"spacy.ner_scorer.v1"}
67
+ update_with_oracle_cut_size = 100
68
+
69
+ [components.ner.model]
70
+ @architectures = "spacy.TransitionBasedParser.v2"
71
+ state_type = "ner"
72
+ extra_state_tokens = false
73
+ hidden_width = 64
74
+ maxout_pieces = 2
75
+ use_upper = true
76
+ nO = null
77
+
78
+ [components.ner.model.tok2vec]
79
+ @architectures = "spacy.Tok2Vec.v2"
80
+
81
+ [components.ner.model.tok2vec.embed]
82
+ @architectures = "spacy.MultiHashEmbed.v2"
83
+ width = 96
84
+ attrs = ["NORM","PREFIX","SUFFIX","SHAPE","SPACY"]
85
+ rows = [5000,2500,2500,2500,100]
86
+ include_static_vectors = true
87
+
88
+ [components.ner.model.tok2vec.encode]
89
+ @architectures = "spacy.MaxoutWindowEncoder.v2"
90
+ width = 96
91
+ depth = 4
92
+ window_size = 1
93
+ maxout_pieces = 3
94
+
95
+ [components.parser]
96
+ factory = "parser"
97
+ learn_tokens = false
98
+ min_action_freq = 30
99
+ moves = null
100
+ scorer = {"@scorers":"spacy.parser_scorer.v1"}
101
+ update_with_oracle_cut_size = 100
102
+
103
+ [components.parser.model]
104
+ @architectures = "spacy.TransitionBasedParser.v2"
105
+ state_type = "parser"
106
+ extra_state_tokens = false
107
+ hidden_width = 64
108
+ maxout_pieces = 2
109
+ use_upper = true
110
+ nO = null
111
+
112
+ [components.parser.model.tok2vec]
113
+ @architectures = "spacy.Tok2Vec.v2"
114
+
115
+ [components.parser.model.tok2vec.embed]
116
+ @architectures = "spacy.MultiHashEmbed.v2"
117
+ width = 96
118
+ attrs = ["NORM","PREFIX","SUFFIX","SHAPE","SPACY"]
119
+ rows = [5000,2500,2500,2500,100]
120
+ include_static_vectors = true
121
+
122
+ [components.parser.model.tok2vec.encode]
123
+ @architectures = "spacy.MaxoutWindowEncoder.v2"
124
+ width = 96
125
+ depth = 4
126
+ window_size = 1
127
+ maxout_pieces = 3
128
+
129
+ [components.tok2vec]
130
+ factory = "tok2vec"
131
+
132
+ [components.tok2vec.model]
133
+ @architectures = "spacy.Tok2Vec.v2"
134
+
135
+ [components.tok2vec.model.embed]
136
+ @architectures = "spacy.MultiHashEmbed.v2"
137
+ width = 96
138
+ attrs = ["NORM","PREFIX","SUFFIX","SHAPE","SPACY"]
139
+ rows = [5000,2500,2500,2500,100]
140
+ include_static_vectors = true
141
+
142
+ [components.tok2vec.model.encode]
143
+ @architectures = "spacy.MaxoutWindowEncoder.v2"
144
+ width = 96
145
+ depth = 4
146
+ window_size = 1
147
+ maxout_pieces = 3
148
+
149
+ [corpora]
150
+ @readers = "prodigy.MergedCorpus.v1"
151
+ eval_split = 0.2
152
+ sample_size = 1.0
153
+ textcat = null
154
+ textcat_multilabel = null
155
+ parser = null
156
+ tagger = null
157
+ senter = null
158
+ spancat = null
159
+
160
+ [corpora.ner]
161
+ @readers = "prodigy.NERCorpus.v1"
162
+ datasets = ["iris_LOC"]
163
+ eval_datasets = []
164
+ default_fill = "outside"
165
+ incorrect_key = "incorrect_spans"
166
+
167
+ [training]
168
+ train_corpus = "corpora.train"
169
+ dev_corpus = "corpora.dev"
170
+ seed = ${system:seed}
171
+ gpu_allocator = ${system:gpu_allocator}
172
+ dropout = 0.1
173
+ accumulate_gradient = 1
174
+ patience = 5000
175
+ max_epochs = 0
176
+ max_steps = 0
177
+ eval_frequency = 1000
178
+ frozen_components = ["morphologizer","parser","attribute_ruler","lemmatizer"]
179
+ before_to_disk = null
180
+ annotating_components = []
181
+
182
+ [training.batcher]
183
+ @batchers = "spacy.batch_by_words.v1"
184
+ discard_oversize = false
185
+ tolerance = 0.2
186
+ get_length = null
187
+
188
+ [training.batcher.size]
189
+ @schedules = "compounding.v1"
190
+ start = 100
191
+ stop = 1000
192
+ compound = 1.001
193
+ t = 0.0
194
+
195
+ [training.logger]
196
+ @loggers = "prodigy.ConsoleLogger.v1"
197
+ progress_bar = false
198
+
199
+ [training.optimizer]
200
+ @optimizers = "Adam.v1"
201
+ beta1 = 0.9
202
+ beta2 = 0.999
203
+ L2_is_weight_decay = true
204
+ L2 = 0.01
205
+ grad_clip = 1.0
206
+ use_averages = true
207
+ eps = 0.00000001
208
+ learn_rate = 0.001
209
+
210
+ [training.score_weights]
211
+ pos_acc = null
212
+ morph_acc = null
213
+ morph_per_feat = null
214
+ dep_uas = null
215
+ dep_las = null
216
+ dep_las_per_type = null
217
+ sents_p = null
218
+ sents_r = null
219
+ sents_f = null
220
+ lemma_acc = null
221
+ ents_f = 1.0
222
+ ents_p = 0.0
223
+ ents_r = 0.0
224
+ ents_per_type = null
225
+ speed = 0.0
226
+
227
+ [pretraining]
228
+
229
+ [initialize]
230
+ vectors = "ca_core_news_lg"
231
+ init_tok2vec = ${paths.init_tok2vec}
232
+ vocab_data = null
233
+ lookups = null
234
+ before_init = null
235
+ after_init = null
236
+
237
+ [initialize.components]
238
+
239
+ [initialize.tokenizer]
lemmatizer/lookups/lookups.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fca90276eaa34797904357bc9070e48e02c586f61331a2c65b1f869fa0025ba
3
+ size 14901088
meta.json ADDED
@@ -0,0 +1,434 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "lang":"ca",
3
+ "name":"anonimization_core_lg",
4
+ "version":"1.0.0",
5
+ "description":"This is a Spacy multilingual anonimization model, for use with BSC's AnonymizationPipeline at https://github.com/TeMU-BSC/AnonymizationPipeline. The anonymization pipeline is a library for performing sensitive data identification and posterior anonymization of the detected data in Spanish and Catalan user generated plain text.",
6
+ "author":"Joaquin Silveira",
7
+ "email":"[email protected]",
8
+ "url":"https://github.com/TeMU-BSC/AnonymizationPipeline",
9
+ "license":"MIT",
10
+ "spacy_version":">=3.2.3,<3.3.0",
11
+ "spacy_git_version":"99425de36",
12
+ "vectors":{
13
+ "width":300,
14
+ "vectors":500000,
15
+ "keys":500000,
16
+ "name":"ca_vectors"
17
+ },
18
+ "labels":{
19
+ "tok2vec":[
20
+
21
+ ],
22
+ "morphologizer":[
23
+ "Definite=Def|Gender=Masc|Number=Sing|POS=DET|PronType=Art",
24
+ "POS=PROPN",
25
+ "POS=PUNCT|PunctSide=Ini|PunctType=Brck",
26
+ "POS=PUNCT|PunctSide=Fin|PunctType=Brck",
27
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin",
28
+ "Gender=Masc|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part",
29
+ "Definite=Def|Gender=Fem|Number=Sing|POS=DET|PronType=Art",
30
+ "Gender=Fem|Number=Sing|POS=NOUN",
31
+ "POS=ADP",
32
+ "NumType=Card|Number=Plur|POS=NUM",
33
+ "Gender=Masc|Number=Plur|POS=NOUN",
34
+ "Number=Sing|POS=ADJ",
35
+ "POS=CCONJ",
36
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Ind",
37
+ "NumForm=Digit|NumType=Card|POS=NUM",
38
+ "NumForm=Digit|POS=NOUN",
39
+ "Gender=Masc|Number=Plur|POS=ADJ",
40
+ "POS=PUNCT|PunctType=Comm",
41
+ "POS=AUX|VerbForm=Inf",
42
+ "Case=Acc,Dat|POS=PRON|Person=3|PrepCase=Npr|PronType=Prs|Reflex=Yes",
43
+ "Definite=Def|Gender=Masc|Number=Plur|POS=DET|PronType=Art",
44
+ "POS=PRON|PronType=Rel",
45
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin",
46
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Art",
47
+ "Gender=Fem|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs",
48
+ "Definite=Def|Gender=Fem|Number=Plur|POS=DET|PronType=Art",
49
+ "Gender=Fem|Number=Plur|POS=NOUN",
50
+ "Gender=Fem|Number=Plur|POS=ADJ",
51
+ "POS=VERB|VerbForm=Inf",
52
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=3|PronType=Prs",
53
+ "Number=Plur|POS=ADJ",
54
+ "POS=PUNCT|PunctType=Peri",
55
+ "Number=Sing|POS=PRON|PronType=Rel",
56
+ "Gender=Masc|Number=Sing|POS=NOUN",
57
+ "Mood=Imp|Number=Sing|POS=VERB|Person=2|VerbForm=Fin",
58
+ "Gender=Masc|Number=Plur|POS=ADJ|VerbForm=Part",
59
+ "POS=SCONJ",
60
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin",
61
+ "Gender=Masc|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part",
62
+ "Definite=Def|Number=Sing|POS=DET|PronType=Art",
63
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Ind",
64
+ "Gender=Fem|Number=Plur|POS=ADJ|VerbForm=Part",
65
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Dem",
66
+ "POS=VERB|VerbForm=Ger",
67
+ "POS=NOUN",
68
+ "Gender=Fem|NumType=Card|Number=Sing|POS=NUM",
69
+ "Gender=Fem|Number=Sing|POS=ADJ|VerbForm=Part",
70
+ "Gender=Fem|NumType=Ord|Number=Plur|POS=ADJ",
71
+ "POS=SYM",
72
+ "Gender=Masc|Number=Sing|POS=ADJ",
73
+ "Gender=Masc|Number=Sing|POS=ADJ|VerbForm=Part",
74
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin",
75
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Dem",
76
+ "POS=ADV|Polarity=Neg",
77
+ "POS=ADV",
78
+ "Number=Sing|POS=PRON|PronType=Dem",
79
+ "Number=Sing|POS=NOUN",
80
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin",
81
+ "Number=Plur|POS=NOUN",
82
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin",
83
+ "Gender=Fem|Number=Sing|POS=ADJ",
84
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin",
85
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Tot",
86
+ "Case=Loc|POS=PRON|Person=3|PronType=Prs",
87
+ "Gender=Fem|NumType=Ord|Number=Sing|POS=ADJ",
88
+ "Degree=Cmp|POS=ADV",
89
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Art",
90
+ "Gender=Fem|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs",
91
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin",
92
+ "Gender=Masc|NumType=Ord|Number=Sing|POS=ADJ",
93
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin",
94
+ "NumType=Card|POS=NUM",
95
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin",
96
+ "Number=Sing|POS=PRON|PronType=Ind",
97
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Art",
98
+ "Number=Plur|POS=DET|PronType=Ind",
99
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin",
100
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Dem",
101
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin",
102
+ "Gender=Masc|NumType=Card|Number=Sing|POS=NUM",
103
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin",
104
+ "Case=Acc|Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs",
105
+ "Number=Sing|POS=DET|PronType=Ind",
106
+ "POS=PUNCT",
107
+ "Number=Sing|POS=DET|PronType=Rel",
108
+ "Case=Gen|POS=PRON|Person=3|PronType=Prs",
109
+ "Gender=Fem|NumType=Card|Number=Plur|POS=NUM",
110
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin",
111
+ "POS=DET|PronType=Ind",
112
+ "POS=AUX",
113
+ "Case=Acc|Gender=Neut|Number=Sing|POS=PRON|Person=3|PronType=Prs",
114
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=1|PronType=Prs",
115
+ "Degree=Cmp|Number=Sing|POS=ADJ",
116
+ "Number=Sing|POS=VERB",
117
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Ind",
118
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Dem",
119
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Art",
120
+ "Gender=Masc|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs",
121
+ "Case=Acc|Gender=Fem,Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs",
122
+ "Gender=Fem|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part",
123
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Ind",
124
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Ind",
125
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin",
126
+ "Number=Plur|POS=PRON|PronType=Rel",
127
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Int",
128
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin",
129
+ "AdvType=Tim|POS=NOUN",
130
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Ind",
131
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Ind",
132
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Int",
133
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=3|VerbForm=Fin",
134
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin",
135
+ "Number=Sing|POS=DET|PronType=Art",
136
+ "Gender=Masc|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs",
137
+ "Case=Acc|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs",
138
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Int",
139
+ "POS=PUNCT|PunctType=Semi",
140
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=3|VerbForm=Fin",
141
+ "Case=Dat|Number=Sing|POS=PRON|Person=3|PronType=Prs",
142
+ "Gender=Masc|NumType=Card|Number=Plur|POS=NUM",
143
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin",
144
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Ind",
145
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin",
146
+ "NumForm=Digit|POS=SYM",
147
+ "Gender=Masc|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part",
148
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Int",
149
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Int",
150
+ "POS=PRON|PronType=Int",
151
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Int",
152
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=3|VerbForm=Fin",
153
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=3|VerbForm=Fin",
154
+ "POS=PART",
155
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Dem",
156
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Tot",
157
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Dem",
158
+ "POS=ADJ",
159
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|PronType=Prs",
160
+ "Degree=Cmp|Number=Plur|POS=ADJ",
161
+ "POS=PUNCT|PunctType=Dash",
162
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin",
163
+ "Case=Acc|Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs",
164
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin",
165
+ "Gender=Fem|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part",
166
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs",
167
+ "Gender=Masc|POS=NOUN",
168
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Past|VerbForm=Fin",
169
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Int",
170
+ "Gender=Masc|NumType=Ord|Number=Plur|POS=ADJ",
171
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin",
172
+ "POS=PUNCT|PunctType=Colo",
173
+ "Gender=Masc|NumType=Card|POS=NUM",
174
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs",
175
+ "Number=Sing|POS=PRON|PronType=Int",
176
+ "POS=PUNCT|PunctType=Quot",
177
+ "Mood=Imp|Number=Sing|POS=VERB|Person=3|VerbForm=Fin",
178
+ "Gender=Fem|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs",
179
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs",
180
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin",
181
+ "POS=AUX|VerbForm=Ger",
182
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs",
183
+ "Mood=Imp|Number=Sing|POS=AUX|Person=3|VerbForm=Fin",
184
+ "Number=Plur|POS=PRON|PronType=Ind",
185
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Dem",
186
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Npr|PronType=Prs",
187
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Int",
188
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin",
189
+ "NumForm=Digit|NumType=Frac|POS=NUM",
190
+ "POS=VERB",
191
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Dem",
192
+ "Gender=Fem|POS=NOUN",
193
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=1|PrepCase=Npr|PronType=Prs",
194
+ "Mood=Sub|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin",
195
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Fut|VerbForm=Fin",
196
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin",
197
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin",
198
+ "Number=Plur|POS=PRON|Person=1|PronType=Prs",
199
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin",
200
+ "Case=Nom|Number=Sing|POS=PRON|Person=2|Polite=Infm|PronType=Prs",
201
+ "POS=X",
202
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=1|VerbForm=Fin",
203
+ "Number=Sing|POS=DET|PronType=Dem",
204
+ "POS=DET",
205
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin",
206
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin",
207
+ "POS=DET|PronType=Art",
208
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs",
209
+ "NumType=Ord|Number=Sing|POS=ADJ",
210
+ "Gender=Fem|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part",
211
+ "Number=Plur|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs",
212
+ "Gender=Fem|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part",
213
+ "Gender=Masc|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part",
214
+ "Number=Plur|POS=PRON|PronType=Dem",
215
+ "Mood=Imp|Number=Plur|POS=VERB|Person=1|VerbForm=Fin",
216
+ "POS=PRON|PronType=Ind",
217
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin",
218
+ "Mood=Imp|Number=Plur|POS=VERB|Person=3|VerbForm=Fin",
219
+ "Case=Nom|Number=Sing|POS=PRON|Person=1|PronType=Prs",
220
+ "Case=Acc|Number=Sing|POS=PRON|Person=1|PrepCase=Pre|PronType=Prs",
221
+ "Mood=Ind|Number=Sing|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin",
222
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin",
223
+ "POS=PUNCT|PunctSide=Fin|PunctType=Qest",
224
+ "NumForm=Digit|NumType=Ord|POS=ADJ",
225
+ "Case=Acc|POS=PRON|Person=3|PrepCase=Pre|PronType=Prs|Reflex=Yes",
226
+ "NumForm=Digit|NumType=Frac|POS=SYM",
227
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin",
228
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs",
229
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs",
230
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin",
231
+ "POS=PUNCT|PunctSide=Ini|PunctType=Qest",
232
+ "NumType=Card|Number=Sing|POS=NUM",
233
+ "Foreign=Yes|POS=PRON|PronType=Int",
234
+ "Foreign=Yes|Mood=Ind|POS=VERB|VerbForm=Fin",
235
+ "Foreign=Yes|POS=ADP",
236
+ "Gender=Masc|Number=Sing|POS=PROPN",
237
+ "POS=PUNCT|PunctSide=Ini|PunctType=Excl",
238
+ "POS=PUNCT|PunctSide=Fin|PunctType=Excl",
239
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=1|VerbForm=Fin",
240
+ "Number=Plur|POS=PRON|Person=2|Polite=Form|PronType=Prs",
241
+ "Mood=Sub|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin",
242
+ "POS=PUNCT|PunctSide=Ini|PunctType=Comm",
243
+ "POS=PUNCT|PunctSide=Fin|PunctType=Comm",
244
+ "Number=Plur|POS=PRON|Person=2|PronType=Prs",
245
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin",
246
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=2|PronType=Prs",
247
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=1|VerbForm=Fin",
248
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=1|VerbForm=Fin",
249
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin",
250
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs",
251
+ "Definite=Ind|Gender=Masc|Number=Sing|POS=DET|PronType=Art",
252
+ "Number=Sing|POS=PRON|Person=2|Polite=Form|PronType=Prs",
253
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs",
254
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin",
255
+ "POS=VERB|Tense=Past|VerbForm=Part",
256
+ "Mood=Imp|Number=Plur|POS=AUX|Person=3|VerbForm=Fin",
257
+ "Case=Nom|POS=PRON|Person=3|PronType=Prs",
258
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Past|VerbForm=Fin",
259
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs",
260
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Rel",
261
+ "Definite=Ind|Number=Sing|POS=DET|PronType=Art",
262
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs",
263
+ "Number=Plur|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs",
264
+ "POS=AUX|Tense=Past|VerbForm=Part",
265
+ "Gender=Fem|NumType=Card|POS=NUM",
266
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin",
267
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin",
268
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs",
269
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin",
270
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Past|VerbForm=Fin",
271
+ "AdvType=Tim|Degree=Cmp|POS=ADV",
272
+ "Case=Acc|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Pre|PronType=Prs",
273
+ "POS=DET|PronType=Rel",
274
+ "Definite=Ind|Gender=Fem|Number=Plur|POS=DET|PronType=Art",
275
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin",
276
+ "POS=INTJ",
277
+ "Mood=Sub|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin",
278
+ "POS=VERB|VerbForm=Fin",
279
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Past|VerbForm=Fin",
280
+ "Definite=Ind|Gender=Fem|Number=Sing|POS=DET|PronType=Art",
281
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin",
282
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs",
283
+ "Mood=Sub|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin",
284
+ "Case=Acc|POS=PRON|Person=3|PronType=Prs|Reflex=Yes",
285
+ "Foreign=Yes|POS=NOUN",
286
+ "Foreign=Yes|Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin",
287
+ "Foreign=Yes|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs",
288
+ "Foreign=Yes|POS=SCONJ",
289
+ "Foreign=Yes|Gender=Fem|Number=Sing|POS=DET|PronType=Art",
290
+ "Gender=Masc|POS=SYM",
291
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs",
292
+ "Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs",
293
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs",
294
+ "Gender=Fem|Number=Sing|POS=PROPN",
295
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin",
296
+ "Definite=Def|Foreign=Yes|Gender=Masc|Number=Sing|POS=DET|PronType=Art",
297
+ "Foreign=Yes|POS=VERB",
298
+ "Foreign=Yes|POS=ADJ",
299
+ "Foreign=Yes|POS=DET",
300
+ "Foreign=Yes|POS=ADV",
301
+ "POS=PUNCT|PunctSide=Fin|Punta d'aignctType=Brck",
302
+ "Degree=Cmp|POS=ADJ",
303
+ "AdvType=Tim|POS=SYM",
304
+ "Number=Plur|POS=DET|PronType=Dem",
305
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin"
306
+ ],
307
+ "parser":[
308
+ "ROOT",
309
+ "acl",
310
+ "advcl",
311
+ "advmod",
312
+ "amod",
313
+ "appos",
314
+ "aux",
315
+ "case",
316
+ "cc",
317
+ "ccomp",
318
+ "compound",
319
+ "conj",
320
+ "cop",
321
+ "csubj",
322
+ "dep",
323
+ "det",
324
+ "expl:pass",
325
+ "fixed",
326
+ "flat",
327
+ "iobj",
328
+ "mark",
329
+ "nmod",
330
+ "nsubj",
331
+ "nummod",
332
+ "obj",
333
+ "obl",
334
+ "parataxis",
335
+ "punct",
336
+ "xcomp"
337
+ ],
338
+ "attribute_ruler":[
339
+
340
+ ],
341
+ "lemmatizer":[
342
+
343
+ ],
344
+ "ner":[
345
+ "EMAIL",
346
+ "FINANCIAL",
347
+ "ID",
348
+ "LOC",
349
+ "MISC",
350
+ "ORG",
351
+ "PER",
352
+ "TELEPHONE",
353
+ "VEHICLE",
354
+ "ZIP"
355
+ ]
356
+ },
357
+ "pipeline":[
358
+ "tok2vec",
359
+ "morphologizer",
360
+ "parser",
361
+ "attribute_ruler",
362
+ "lemmatizer",
363
+ "ner"
364
+ ],
365
+ "components":[
366
+ "tok2vec",
367
+ "morphologizer",
368
+ "parser",
369
+ "attribute_ruler",
370
+ "lemmatizer",
371
+ "ner"
372
+ ],
373
+ "disabled":[
374
+
375
+ ],
376
+ "performance":{
377
+ "pos_acc":0.0,
378
+ "morph_acc":0.0,
379
+ "morph_per_feat":0.0,
380
+ "dep_uas":0.0,
381
+ "dep_las":0.0,
382
+ "dep_las_per_type":0.0,
383
+ "sents_p":0.0,
384
+ "sents_r":0.0,
385
+ "sents_f":0.0,
386
+ "lemma_acc":0.0,
387
+ "ents_f":0.6911764706,
388
+ "ents_p":0.746031746,
389
+ "ents_r":0.6438356164,
390
+ "ents_per_type":{
391
+ "LOC":{
392
+ "p":0.652892562,
393
+ "r":0.5895522388,
394
+ "f":0.6196078431
395
+ },
396
+ "ID":{
397
+ "p":0.8,
398
+ "r":0.5714285714,
399
+ "f":0.6666666667
400
+ },
401
+ "PER":{
402
+ "p":0.8695652174,
403
+ "r":0.625,
404
+ "f":0.7272727273
405
+ },
406
+ "EMAIL":{
407
+ "p":0.8571428571,
408
+ "r":1.0,
409
+ "f":0.9230769231
410
+ },
411
+ "ZIP":{
412
+ "p":1.0,
413
+ "r":1.0,
414
+ "f":1.0
415
+ },
416
+ "TELEPHONE":{
417
+ "p":1.0,
418
+ "r":0.8888888889,
419
+ "f":0.9411764706
420
+ },
421
+ "VEHICLE":{
422
+ "p":1.0,
423
+ "r":0.7894736842,
424
+ "f":0.8823529412
425
+ }
426
+ },
427
+ "speed":7904.1995709831,
428
+ "tok2vec_loss":0.0,
429
+ "ner_loss":265.7377590023
430
+ },
431
+ "requirements":[
432
+
433
+ ]
434
+ }
morphologizer/cfg ADDED
@@ -0,0 +1,574 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "extend":false,
3
+ "labels_morph":{
4
+ "Definite=Def|Gender=Masc|Number=Sing|POS=DET|PronType=Art":"Definite=Def|Gender=Masc|Number=Sing|PronType=Art",
5
+ "POS=PROPN":"",
6
+ "POS=PUNCT|PunctSide=Ini|PunctType=Brck":"PunctSide=Ini|PunctType=Brck",
7
+ "POS=PUNCT|PunctSide=Fin|PunctType=Brck":"PunctSide=Fin|PunctType=Brck",
8
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin",
9
+ "Gender=Masc|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part":"Gender=Masc|Number=Sing|Tense=Past|VerbForm=Part",
10
+ "Definite=Def|Gender=Fem|Number=Sing|POS=DET|PronType=Art":"Definite=Def|Gender=Fem|Number=Sing|PronType=Art",
11
+ "Gender=Fem|Number=Sing|POS=NOUN":"Gender=Fem|Number=Sing",
12
+ "POS=ADP":"",
13
+ "NumType=Card|Number=Plur|POS=NUM":"NumType=Card|Number=Plur",
14
+ "Gender=Masc|Number=Plur|POS=NOUN":"Gender=Masc|Number=Plur",
15
+ "Number=Sing|POS=ADJ":"Number=Sing",
16
+ "POS=CCONJ":"",
17
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Ind":"Gender=Fem|Number=Sing|PronType=Ind",
18
+ "NumForm=Digit|NumType=Card|POS=NUM":"NumForm=Digit|NumType=Card",
19
+ "NumForm=Digit|POS=NOUN":"NumForm=Digit",
20
+ "Gender=Masc|Number=Plur|POS=ADJ":"Gender=Masc|Number=Plur",
21
+ "POS=PUNCT|PunctType=Comm":"PunctType=Comm",
22
+ "POS=AUX|VerbForm=Inf":"VerbForm=Inf",
23
+ "Case=Acc,Dat|POS=PRON|Person=3|PrepCase=Npr|PronType=Prs|Reflex=Yes":"Case=Acc,Dat|Person=3|PrepCase=Npr|PronType=Prs|Reflex=Yes",
24
+ "Definite=Def|Gender=Masc|Number=Plur|POS=DET|PronType=Art":"Definite=Def|Gender=Masc|Number=Plur|PronType=Art",
25
+ "POS=PRON|PronType=Rel":"PronType=Rel",
26
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Imp|VerbForm=Fin",
27
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Art":"Gender=Fem|Number=Sing|PronType=Art",
28
+ "Gender=Fem|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Sing|Person=3|Poss=Yes|PronType=Prs",
29
+ "Definite=Def|Gender=Fem|Number=Plur|POS=DET|PronType=Art":"Definite=Def|Gender=Fem|Number=Plur|PronType=Art",
30
+ "Gender=Fem|Number=Plur|POS=NOUN":"Gender=Fem|Number=Plur",
31
+ "Gender=Fem|Number=Plur|POS=ADJ":"Gender=Fem|Number=Plur",
32
+ "POS=VERB|VerbForm=Inf":"VerbForm=Inf",
33
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=3|PronType=Prs":"Case=Acc,Dat|Number=Plur|Person=3|PronType=Prs",
34
+ "Number=Plur|POS=ADJ":"Number=Plur",
35
+ "POS=PUNCT|PunctType=Peri":"PunctType=Peri",
36
+ "Number=Sing|POS=PRON|PronType=Rel":"Number=Sing|PronType=Rel",
37
+ "Gender=Masc|Number=Sing|POS=NOUN":"Gender=Masc|Number=Sing",
38
+ "Mood=Imp|Number=Sing|POS=VERB|Person=2|VerbForm=Fin":"Mood=Imp|Number=Sing|Person=2|VerbForm=Fin",
39
+ "Gender=Masc|Number=Plur|POS=ADJ|VerbForm=Part":"Gender=Masc|Number=Plur|VerbForm=Part",
40
+ "POS=SCONJ":"",
41
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Pres|VerbForm=Fin",
42
+ "Gender=Masc|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part":"Gender=Masc|Number=Plur|Tense=Past|VerbForm=Part",
43
+ "Definite=Def|Number=Sing|POS=DET|PronType=Art":"Definite=Def|Number=Sing|PronType=Art",
44
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Ind":"Gender=Masc|Number=Sing|PronType=Ind",
45
+ "Gender=Fem|Number=Plur|POS=ADJ|VerbForm=Part":"Gender=Fem|Number=Plur|VerbForm=Part",
46
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Dem":"Gender=Masc|Number=Sing|PronType=Dem",
47
+ "POS=VERB|VerbForm=Ger":"VerbForm=Ger",
48
+ "POS=NOUN":"",
49
+ "Gender=Fem|NumType=Card|Number=Sing|POS=NUM":"Gender=Fem|NumType=Card|Number=Sing",
50
+ "Gender=Fem|Number=Sing|POS=ADJ|VerbForm=Part":"Gender=Fem|Number=Sing|VerbForm=Part",
51
+ "Gender=Fem|NumType=Ord|Number=Plur|POS=ADJ":"Gender=Fem|NumType=Ord|Number=Plur",
52
+ "POS=SYM":"",
53
+ "Gender=Masc|Number=Sing|POS=ADJ":"Gender=Masc|Number=Sing",
54
+ "Gender=Masc|Number=Sing|POS=ADJ|VerbForm=Part":"Gender=Masc|Number=Sing|VerbForm=Part",
55
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin",
56
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Dem":"Gender=Fem|Number=Sing|PronType=Dem",
57
+ "POS=ADV|Polarity=Neg":"Polarity=Neg",
58
+ "POS=ADV":"",
59
+ "Number=Sing|POS=PRON|PronType=Dem":"Number=Sing|PronType=Dem",
60
+ "Number=Sing|POS=NOUN":"Number=Sing",
61
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Pres|VerbForm=Fin",
62
+ "Number=Plur|POS=NOUN":"Number=Plur",
63
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=3|Tense=Imp|VerbForm=Fin",
64
+ "Gender=Fem|Number=Sing|POS=ADJ":"Gender=Fem|Number=Sing",
65
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=1|Tense=Pres|VerbForm=Fin",
66
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Tot":"Gender=Masc|Number=Sing|PronType=Tot",
67
+ "Case=Loc|POS=PRON|Person=3|PronType=Prs":"Case=Loc|Person=3|PronType=Prs",
68
+ "Gender=Fem|NumType=Ord|Number=Sing|POS=ADJ":"Gender=Fem|NumType=Ord|Number=Sing",
69
+ "Degree=Cmp|POS=ADV":"Degree=Cmp",
70
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Art":"Gender=Fem|Number=Plur|PronType=Art",
71
+ "Gender=Fem|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Plur|Person=3|Poss=Yes|PronType=Prs",
72
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Fut|VerbForm=Fin",
73
+ "Gender=Masc|NumType=Ord|Number=Sing|POS=ADJ":"Gender=Masc|NumType=Ord|Number=Sing",
74
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Fut|VerbForm=Fin",
75
+ "NumType=Card|POS=NUM":"NumType=Card",
76
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Fut|VerbForm=Fin",
77
+ "Number=Sing|POS=PRON|PronType=Ind":"Number=Sing|PronType=Ind",
78
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Art":"Gender=Masc|Number=Sing|PronType=Art",
79
+ "Number=Plur|POS=DET|PronType=Ind":"Number=Plur|PronType=Ind",
80
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=3|Tense=Pres|VerbForm=Fin",
81
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Dem":"Gender=Masc|Number=Plur|PronType=Dem",
82
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Fut|VerbForm=Fin",
83
+ "Gender=Masc|NumType=Card|Number=Sing|POS=NUM":"Gender=Masc|NumType=Card|Number=Sing",
84
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=3|Tense=Pres|VerbForm=Fin",
85
+ "Case=Acc|Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Case=Acc|Gender=Fem|Number=Sing|Person=3|PronType=Prs",
86
+ "Number=Sing|POS=DET|PronType=Ind":"Number=Sing|PronType=Ind",
87
+ "POS=PUNCT":"",
88
+ "Number=Sing|POS=DET|PronType=Rel":"Number=Sing|PronType=Rel",
89
+ "Case=Gen|POS=PRON|Person=3|PronType=Prs":"Case=Gen|Person=3|PronType=Prs",
90
+ "Gender=Fem|NumType=Card|Number=Plur|POS=NUM":"Gender=Fem|NumType=Card|Number=Plur",
91
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Pres|VerbForm=Fin",
92
+ "POS=DET|PronType=Ind":"PronType=Ind",
93
+ "POS=AUX":"",
94
+ "Case=Acc|Gender=Neut|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Case=Acc|Gender=Neut|Number=Sing|Person=3|PronType=Prs",
95
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=1|PronType=Prs":"Case=Acc,Dat|Number=Plur|Person=1|PronType=Prs",
96
+ "Degree=Cmp|Number=Sing|POS=ADJ":"Degree=Cmp|Number=Sing",
97
+ "Number=Sing|POS=VERB":"Number=Sing",
98
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Ind":"Gender=Masc|Number=Plur|PronType=Ind",
99
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Dem":"Gender=Fem|Number=Plur|PronType=Dem",
100
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Art":"Gender=Masc|Number=Plur|PronType=Art",
101
+ "Gender=Masc|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Plur|Person=3|Poss=Yes|PronType=Prs",
102
+ "Case=Acc|Gender=Fem,Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Case=Acc|Gender=Fem,Masc|Number=Sing|Person=3|PronType=Prs",
103
+ "Gender=Fem|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part":"Gender=Fem|Number=Sing|Tense=Past|VerbForm=Part",
104
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Ind":"Gender=Masc|Number=Sing|PronType=Ind",
105
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Ind":"Gender=Fem|Number=Plur|PronType=Ind",
106
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin",
107
+ "Number=Plur|POS=PRON|PronType=Rel":"Number=Plur|PronType=Rel",
108
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Int":"Gender=Masc|Number=Plur|PronType=Int",
109
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Imp|VerbForm=Fin",
110
+ "AdvType=Tim|POS=NOUN":"AdvType=Tim",
111
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Ind":"Gender=Masc|Number=Plur|PronType=Ind",
112
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Ind":"Gender=Fem|Number=Plur|PronType=Ind",
113
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Int":"Gender=Masc|Number=Sing|PronType=Int",
114
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=3|VerbForm=Fin":"Mood=Cnd|Number=Sing|Person=3|VerbForm=Fin",
115
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Imp|VerbForm=Fin",
116
+ "Number=Sing|POS=DET|PronType=Art":"Number=Sing|PronType=Art",
117
+ "Gender=Masc|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Person=3|Poss=Yes|PronType=Prs",
118
+ "Case=Acc|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Case=Acc|Gender=Masc|Number=Sing|Person=3|PronType=Prs",
119
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Int":"Gender=Masc|Number=Sing|PronType=Int",
120
+ "POS=PUNCT|PunctType=Semi":"PunctType=Semi",
121
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=3|VerbForm=Fin":"Mood=Cnd|Number=Plur|Person=3|VerbForm=Fin",
122
+ "Case=Dat|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Case=Dat|Number=Sing|Person=3|PronType=Prs",
123
+ "Gender=Masc|NumType=Card|Number=Plur|POS=NUM":"Gender=Masc|NumType=Card|Number=Plur",
124
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Imp|VerbForm=Fin",
125
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Ind":"Gender=Fem|Number=Sing|PronType=Ind",
126
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=3|Tense=Imp|VerbForm=Fin",
127
+ "NumForm=Digit|POS=SYM":"NumForm=Digit",
128
+ "Gender=Masc|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part":"Gender=Masc|Number=Sing|Tense=Past|VerbForm=Part",
129
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Int":"Gender=Fem|Number=Sing|PronType=Int",
130
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Int":"Gender=Fem|Number=Sing|PronType=Int",
131
+ "POS=PRON|PronType=Int":"PronType=Int",
132
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Int":"Gender=Fem|Number=Plur|PronType=Int",
133
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=3|VerbForm=Fin":"Mood=Cnd|Number=Sing|Person=3|VerbForm=Fin",
134
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=3|VerbForm=Fin":"Mood=Cnd|Number=Plur|Person=3|VerbForm=Fin",
135
+ "POS=PART":"",
136
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Dem":"Gender=Fem|Number=Sing|PronType=Dem",
137
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Tot":"Gender=Masc|Number=Sing|PronType=Tot",
138
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Dem":"Gender=Masc|Number=Plur|PronType=Dem",
139
+ "POS=ADJ":"",
140
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|PronType=Prs":"Gender=Masc|Number=Plur|Person=3|PronType=Prs",
141
+ "Degree=Cmp|Number=Plur|POS=ADJ":"Degree=Cmp|Number=Plur",
142
+ "POS=PUNCT|PunctType=Dash":"PunctType=Dash",
143
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin",
144
+ "Case=Acc|Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs":"Case=Acc|Gender=Fem|Number=Plur|Person=3|PronType=Prs",
145
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=3|Tense=Imp|VerbForm=Fin",
146
+ "Gender=Fem|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part":"Gender=Fem|Number=Plur|Tense=Past|VerbForm=Part",
147
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Gender=Fem|Number=Sing|Person=3|PronType=Prs",
148
+ "Gender=Masc|POS=NOUN":"Gender=Masc",
149
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Past|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Past|VerbForm=Fin",
150
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Int":"Gender=Fem|Number=Plur|PronType=Int",
151
+ "Gender=Masc|NumType=Ord|Number=Plur|POS=ADJ":"Gender=Masc|NumType=Ord|Number=Plur",
152
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Fut|VerbForm=Fin",
153
+ "POS=PUNCT|PunctType=Colo":"PunctType=Colo",
154
+ "Gender=Masc|NumType=Card|POS=NUM":"Gender=Masc|NumType=Card",
155
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Gender=Masc|Number=Sing|Person=3|PronType=Prs",
156
+ "Number=Sing|POS=PRON|PronType=Int":"Number=Sing|PronType=Int",
157
+ "POS=PUNCT|PunctType=Quot":"PunctType=Quot",
158
+ "Mood=Imp|Number=Sing|POS=VERB|Person=3|VerbForm=Fin":"Mood=Imp|Number=Sing|Person=3|VerbForm=Fin",
159
+ "Gender=Fem|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs",
160
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs",
161
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Fut|VerbForm=Fin",
162
+ "POS=AUX|VerbForm=Ger":"VerbForm=Ger",
163
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs":"Gender=Fem|Number=Plur|Person=3|PronType=Prs",
164
+ "Mood=Imp|Number=Sing|POS=AUX|Person=3|VerbForm=Fin":"Mood=Imp|Number=Sing|Person=3|VerbForm=Fin",
165
+ "Number=Plur|POS=PRON|PronType=Ind":"Number=Plur|PronType=Ind",
166
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Dem":"Gender=Masc|Number=Sing|PronType=Dem",
167
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Npr|PronType=Prs":"Case=Acc,Dat|Number=Sing|Person=2|Polite=Infm|PrepCase=Npr|PronType=Prs",
168
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Int":"Gender=Masc|Number=Plur|PronType=Int",
169
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Pres|VerbForm=Fin",
170
+ "NumForm=Digit|NumType=Frac|POS=NUM":"NumForm=Digit|NumType=Frac",
171
+ "POS=VERB":"",
172
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Dem":"Gender=Fem|Number=Plur|PronType=Dem",
173
+ "Gender=Fem|POS=NOUN":"Gender=Fem",
174
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=1|PrepCase=Npr|PronType=Prs":"Case=Acc,Dat|Number=Sing|Person=1|PrepCase=Npr|PronType=Prs",
175
+ "Mood=Sub|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=2|Tense=Pres|VerbForm=Fin",
176
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=2|Tense=Fut|VerbForm=Fin",
177
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=1|Tense=Pres|VerbForm=Fin",
178
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=3|Tense=Imp|VerbForm=Fin",
179
+ "Number=Plur|POS=PRON|Person=1|PronType=Prs":"Number=Plur|Person=1|PronType=Prs",
180
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Pres|VerbForm=Fin",
181
+ "Case=Nom|Number=Sing|POS=PRON|Person=2|Polite=Infm|PronType=Prs":"Case=Nom|Number=Sing|Person=2|Polite=Infm|PronType=Prs",
182
+ "POS=X":"",
183
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=1|VerbForm=Fin":"Mood=Cnd|Number=Plur|Person=1|VerbForm=Fin",
184
+ "Number=Sing|POS=DET|PronType=Dem":"Number=Sing|PronType=Dem",
185
+ "POS=DET":"",
186
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Fut|VerbForm=Fin",
187
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Pres|VerbForm=Fin",
188
+ "POS=DET|PronType=Art":"PronType=Art",
189
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Person=3|Poss=Yes|PronType=Prs",
190
+ "NumType=Ord|Number=Sing|POS=ADJ":"NumType=Ord|Number=Sing",
191
+ "Gender=Fem|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part":"Gender=Fem|Number=Sing|Tense=Past|VerbForm=Part",
192
+ "Number=Plur|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":"Number=Plur|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs",
193
+ "Gender=Fem|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part":"Gender=Fem|Number=Plur|Tense=Past|VerbForm=Part",
194
+ "Gender=Masc|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part":"Gender=Masc|Number=Plur|Tense=Past|VerbForm=Part",
195
+ "Number=Plur|POS=PRON|PronType=Dem":"Number=Plur|PronType=Dem",
196
+ "Mood=Imp|Number=Plur|POS=VERB|Person=1|VerbForm=Fin":"Mood=Imp|Number=Plur|Person=1|VerbForm=Fin",
197
+ "POS=PRON|PronType=Ind":"PronType=Ind",
198
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=2|Tense=Pres|VerbForm=Fin",
199
+ "Mood=Imp|Number=Plur|POS=VERB|Person=3|VerbForm=Fin":"Mood=Imp|Number=Plur|Person=3|VerbForm=Fin",
200
+ "Case=Nom|Number=Sing|POS=PRON|Person=1|PronType=Prs":"Case=Nom|Number=Sing|Person=1|PronType=Prs",
201
+ "Case=Acc|Number=Sing|POS=PRON|Person=1|PrepCase=Pre|PronType=Prs":"Case=Acc|Number=Sing|Person=1|PrepCase=Pre|PronType=Prs",
202
+ "Mood=Ind|Number=Sing|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=2|Tense=Pres|VerbForm=Fin",
203
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Imp|VerbForm=Fin",
204
+ "POS=PUNCT|PunctSide=Fin|PunctType=Qest":"PunctSide=Fin|PunctType=Qest",
205
+ "NumForm=Digit|NumType=Ord|POS=ADJ":"NumForm=Digit|NumType=Ord",
206
+ "Case=Acc|POS=PRON|Person=3|PrepCase=Pre|PronType=Prs|Reflex=Yes":"Case=Acc|Person=3|PrepCase=Pre|PronType=Prs|Reflex=Yes",
207
+ "NumForm=Digit|NumType=Frac|POS=SYM":"NumForm=Digit|NumType=Frac",
208
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=2|Tense=Pres|VerbForm=Fin",
209
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs",
210
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Plur|Person=3|Poss=Yes|PronType=Prs",
211
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=1|Tense=Pres|VerbForm=Fin",
212
+ "POS=PUNCT|PunctSide=Ini|PunctType=Qest":"PunctSide=Ini|PunctType=Qest",
213
+ "NumType=Card|Number=Sing|POS=NUM":"NumType=Card|Number=Sing",
214
+ "Foreign=Yes|POS=PRON|PronType=Int":"Foreign=Yes|PronType=Int",
215
+ "Foreign=Yes|Mood=Ind|POS=VERB|VerbForm=Fin":"Foreign=Yes|Mood=Ind|VerbForm=Fin",
216
+ "Foreign=Yes|POS=ADP":"Foreign=Yes",
217
+ "Gender=Masc|Number=Sing|POS=PROPN":"Gender=Masc|Number=Sing",
218
+ "POS=PUNCT|PunctSide=Ini|PunctType=Excl":"PunctSide=Ini|PunctType=Excl",
219
+ "POS=PUNCT|PunctSide=Fin|PunctType=Excl":"PunctSide=Fin|PunctType=Excl",
220
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=1|VerbForm=Fin":"Mood=Cnd|Number=Sing|Person=1|VerbForm=Fin",
221
+ "Number=Plur|POS=PRON|Person=2|Polite=Form|PronType=Prs":"Number=Plur|Person=2|Polite=Form|PronType=Prs",
222
+ "Mood=Sub|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Sub|Person=1|Tense=Imp|VerbForm=Fin",
223
+ "POS=PUNCT|PunctSide=Ini|PunctType=Comm":"PunctSide=Ini|PunctType=Comm",
224
+ "POS=PUNCT|PunctSide=Fin|PunctType=Comm":"PunctSide=Fin|PunctType=Comm",
225
+ "Number=Plur|POS=PRON|Person=2|PronType=Prs":"Number=Plur|Person=2|PronType=Prs",
226
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=2|Tense=Pres|VerbForm=Fin",
227
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=2|PronType=Prs":"Case=Acc,Dat|Number=Plur|Person=2|PronType=Prs",
228
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=1|VerbForm=Fin":"Mood=Cnd|Number=Sing|Person=1|VerbForm=Fin",
229
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=1|VerbForm=Fin":"Mood=Cnd|Number=Plur|Person=1|VerbForm=Fin",
230
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=1|Tense=Imp|VerbForm=Fin",
231
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Plur|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs",
232
+ "Definite=Ind|Gender=Masc|Number=Sing|POS=DET|PronType=Art":"Definite=Ind|Gender=Masc|Number=Sing|PronType=Art",
233
+ "Number=Sing|POS=PRON|Person=2|Polite=Form|PronType=Prs":"Number=Sing|Person=2|Polite=Form|PronType=Prs",
234
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Number[psor]=Sing|Person=1|Poss=Yes|PronType=Prs",
235
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Imp|VerbForm=Fin",
236
+ "POS=VERB|Tense=Past|VerbForm=Part":"Tense=Past|VerbForm=Part",
237
+ "Mood=Imp|Number=Plur|POS=AUX|Person=3|VerbForm=Fin":"Mood=Imp|Number=Plur|Person=3|VerbForm=Fin",
238
+ "Case=Nom|POS=PRON|Person=3|PronType=Prs":"Case=Nom|Person=3|PronType=Prs",
239
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Past|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=3|Tense=Past|VerbForm=Fin",
240
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Sing|Person=3|Poss=Yes|PronType=Prs",
241
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Rel":"Gender=Masc|Number=Sing|PronType=Rel",
242
+ "Definite=Ind|Number=Sing|POS=DET|PronType=Art":"Definite=Ind|Number=Sing|PronType=Art",
243
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Sing|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs",
244
+ "Number=Plur|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs":"Number=Plur|Number[psor]=Plur|Person=1|Poss=Yes|PronType=Prs",
245
+ "POS=AUX|Tense=Past|VerbForm=Part":"Tense=Past|VerbForm=Part",
246
+ "Gender=Fem|NumType=Card|POS=NUM":"Gender=Fem|NumType=Card",
247
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Imp|VerbForm=Fin",
248
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=1|Tense=Imp|VerbForm=Fin",
249
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Plur|Person=3|Poss=Yes|PronType=Prs",
250
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=1|Tense=Fut|VerbForm=Fin",
251
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Past|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Past|VerbForm=Fin",
252
+ "AdvType=Tim|Degree=Cmp|POS=ADV":"AdvType=Tim|Degree=Cmp",
253
+ "Case=Acc|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Pre|PronType=Prs":"Case=Acc|Number=Sing|Person=2|Polite=Infm|PrepCase=Pre|PronType=Prs",
254
+ "POS=DET|PronType=Rel":"PronType=Rel",
255
+ "Definite=Ind|Gender=Fem|Number=Plur|POS=DET|PronType=Art":"Definite=Ind|Gender=Fem|Number=Plur|PronType=Art",
256
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=2|Tense=Fut|VerbForm=Fin",
257
+ "POS=INTJ":"",
258
+ "Mood=Sub|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=1|Tense=Pres|VerbForm=Fin",
259
+ "POS=VERB|VerbForm=Fin":"VerbForm=Fin",
260
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Past|VerbForm=Fin":"Mood=Ind|Number=Plur|Person=3|Tense=Past|VerbForm=Fin",
261
+ "Definite=Ind|Gender=Fem|Number=Sing|POS=DET|PronType=Art":"Definite=Ind|Gender=Fem|Number=Sing|PronType=Art",
262
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=1|Tense=Imp|VerbForm=Fin",
263
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Poss=Yes|PronType=Prs",
264
+ "Mood=Sub|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":"Mood=Sub|Number=Sing|Person=2|Tense=Pres|VerbForm=Fin",
265
+ "Case=Acc|POS=PRON|Person=3|PronType=Prs|Reflex=Yes":"Case=Acc|Person=3|PronType=Prs|Reflex=Yes",
266
+ "Foreign=Yes|POS=NOUN":"Foreign=Yes",
267
+ "Foreign=Yes|Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":"Foreign=Yes|Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin",
268
+ "Foreign=Yes|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":"Foreign=Yes|Gender=Masc|Number=Sing|Person=3|PronType=Prs",
269
+ "Foreign=Yes|POS=SCONJ":"Foreign=Yes",
270
+ "Foreign=Yes|Gender=Fem|Number=Sing|POS=DET|PronType=Art":"Foreign=Yes|Gender=Fem|Number=Sing|PronType=Art",
271
+ "Gender=Masc|POS=SYM":"Gender=Masc",
272
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":"Gender=Fem|Number=Sing|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs",
273
+ "Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":"Number=Sing|Person=3|Poss=Yes|PronType=Prs",
274
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":"Gender=Masc|Number=Plur|Number[psor]=Sing|Person=2|Poss=Yes|PronType=Prs",
275
+ "Gender=Fem|Number=Sing|POS=PROPN":"Gender=Fem|Number=Sing",
276
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":"Mood=Sub|Number=Plur|Person=1|Tense=Imp|VerbForm=Fin",
277
+ "Definite=Def|Foreign=Yes|Gender=Masc|Number=Sing|POS=DET|PronType=Art":"Definite=Def|Foreign=Yes|Gender=Masc|Number=Sing|PronType=Art",
278
+ "Foreign=Yes|POS=VERB":"Foreign=Yes",
279
+ "Foreign=Yes|POS=ADJ":"Foreign=Yes",
280
+ "Foreign=Yes|POS=DET":"Foreign=Yes",
281
+ "Foreign=Yes|POS=ADV":"Foreign=Yes",
282
+ "POS=PUNCT|PunctSide=Fin|Punta d'aignctType=Brck":"PunctSide=Fin|Punta d'aignctType=Brck",
283
+ "Degree=Cmp|POS=ADJ":"Degree=Cmp",
284
+ "AdvType=Tim|POS=SYM":"AdvType=Tim",
285
+ "Number=Plur|POS=DET|PronType=Dem":"Number=Plur|PronType=Dem",
286
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin":"Mood=Ind|Number=Sing|Person=2|Tense=Fut|VerbForm=Fin"
287
+ },
288
+ "labels_pos":{
289
+ "Definite=Def|Gender=Masc|Number=Sing|POS=DET|PronType=Art":90,
290
+ "POS=PROPN":96,
291
+ "POS=PUNCT|PunctSide=Ini|PunctType=Brck":97,
292
+ "POS=PUNCT|PunctSide=Fin|PunctType=Brck":97,
293
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":87,
294
+ "Gender=Masc|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part":100,
295
+ "Definite=Def|Gender=Fem|Number=Sing|POS=DET|PronType=Art":90,
296
+ "Gender=Fem|Number=Sing|POS=NOUN":92,
297
+ "POS=ADP":85,
298
+ "NumType=Card|Number=Plur|POS=NUM":93,
299
+ "Gender=Masc|Number=Plur|POS=NOUN":92,
300
+ "Number=Sing|POS=ADJ":84,
301
+ "POS=CCONJ":89,
302
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Ind":90,
303
+ "NumForm=Digit|NumType=Card|POS=NUM":93,
304
+ "NumForm=Digit|POS=NOUN":92,
305
+ "Gender=Masc|Number=Plur|POS=ADJ":84,
306
+ "POS=PUNCT|PunctType=Comm":97,
307
+ "POS=AUX|VerbForm=Inf":87,
308
+ "Case=Acc,Dat|POS=PRON|Person=3|PrepCase=Npr|PronType=Prs|Reflex=Yes":95,
309
+ "Definite=Def|Gender=Masc|Number=Plur|POS=DET|PronType=Art":90,
310
+ "POS=PRON|PronType=Rel":95,
311
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":100,
312
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Art":90,
313
+ "Gender=Fem|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":90,
314
+ "Definite=Def|Gender=Fem|Number=Plur|POS=DET|PronType=Art":90,
315
+ "Gender=Fem|Number=Plur|POS=NOUN":92,
316
+ "Gender=Fem|Number=Plur|POS=ADJ":84,
317
+ "POS=VERB|VerbForm=Inf":100,
318
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=3|PronType=Prs":95,
319
+ "Number=Plur|POS=ADJ":84,
320
+ "POS=PUNCT|PunctType=Peri":97,
321
+ "Number=Sing|POS=PRON|PronType=Rel":95,
322
+ "Gender=Masc|Number=Sing|POS=NOUN":92,
323
+ "Mood=Imp|Number=Sing|POS=VERB|Person=2|VerbForm=Fin":100,
324
+ "Gender=Masc|Number=Plur|POS=ADJ|VerbForm=Part":84,
325
+ "POS=SCONJ":98,
326
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":87,
327
+ "Gender=Masc|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part":100,
328
+ "Definite=Def|Number=Sing|POS=DET|PronType=Art":90,
329
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Ind":90,
330
+ "Gender=Fem|Number=Plur|POS=ADJ|VerbForm=Part":84,
331
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Dem":90,
332
+ "POS=VERB|VerbForm=Ger":100,
333
+ "POS=NOUN":92,
334
+ "Gender=Fem|NumType=Card|Number=Sing|POS=NUM":93,
335
+ "Gender=Fem|Number=Sing|POS=ADJ|VerbForm=Part":84,
336
+ "Gender=Fem|NumType=Ord|Number=Plur|POS=ADJ":84,
337
+ "POS=SYM":99,
338
+ "Gender=Masc|Number=Sing|POS=ADJ":84,
339
+ "Gender=Masc|Number=Sing|POS=ADJ|VerbForm=Part":84,
340
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":100,
341
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Dem":90,
342
+ "POS=ADV|Polarity=Neg":86,
343
+ "POS=ADV":86,
344
+ "Number=Sing|POS=PRON|PronType=Dem":95,
345
+ "Number=Sing|POS=NOUN":92,
346
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":100,
347
+ "Number=Plur|POS=NOUN":92,
348
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":100,
349
+ "Gender=Fem|Number=Sing|POS=ADJ":84,
350
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":100,
351
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Tot":95,
352
+ "Case=Loc|POS=PRON|Person=3|PronType=Prs":95,
353
+ "Gender=Fem|NumType=Ord|Number=Sing|POS=ADJ":84,
354
+ "Degree=Cmp|POS=ADV":86,
355
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Art":90,
356
+ "Gender=Fem|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs":90,
357
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin":100,
358
+ "Gender=Masc|NumType=Ord|Number=Sing|POS=ADJ":84,
359
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin":87,
360
+ "NumType=Card|POS=NUM":93,
361
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Fut|VerbForm=Fin":100,
362
+ "Number=Sing|POS=PRON|PronType=Ind":95,
363
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Art":90,
364
+ "Number=Plur|POS=DET|PronType=Ind":90,
365
+ "Mood=Sub|Number=Plur|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":100,
366
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Dem":90,
367
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Fut|VerbForm=Fin":87,
368
+ "Gender=Masc|NumType=Card|Number=Sing|POS=NUM":93,
369
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":87,
370
+ "Case=Acc|Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
371
+ "Number=Sing|POS=DET|PronType=Ind":90,
372
+ "POS=PUNCT":97,
373
+ "Number=Sing|POS=DET|PronType=Rel":90,
374
+ "Case=Gen|POS=PRON|Person=3|PronType=Prs":95,
375
+ "Gender=Fem|NumType=Card|Number=Plur|POS=NUM":93,
376
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":100,
377
+ "POS=DET|PronType=Ind":90,
378
+ "POS=AUX":87,
379
+ "Case=Acc|Gender=Neut|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
380
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=1|PronType=Prs":95,
381
+ "Degree=Cmp|Number=Sing|POS=ADJ":84,
382
+ "Number=Sing|POS=VERB":100,
383
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Ind":95,
384
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Dem":90,
385
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Art":90,
386
+ "Gender=Masc|Number=Plur|POS=DET|Person=3|Poss=Yes|PronType=Prs":90,
387
+ "Case=Acc|Gender=Fem,Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
388
+ "Gender=Fem|Number=Sing|POS=VERB|Tense=Past|VerbForm=Part":100,
389
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Ind":95,
390
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Ind":95,
391
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Pres|VerbForm=Fin":100,
392
+ "Number=Plur|POS=PRON|PronType=Rel":95,
393
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Int":90,
394
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":87,
395
+ "AdvType=Tim|POS=NOUN":92,
396
+ "Gender=Masc|Number=Plur|POS=DET|PronType=Ind":90,
397
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Ind":90,
398
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Int":90,
399
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=3|VerbForm=Fin":87,
400
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":100,
401
+ "Number=Sing|POS=DET|PronType=Art":90,
402
+ "Gender=Masc|Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":90,
403
+ "Case=Acc|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
404
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Int":95,
405
+ "POS=PUNCT|PunctType=Semi":97,
406
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=3|VerbForm=Fin":87,
407
+ "Case=Dat|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
408
+ "Gender=Masc|NumType=Card|Number=Plur|POS=NUM":93,
409
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":87,
410
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Ind":95,
411
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":87,
412
+ "NumForm=Digit|POS=SYM":99,
413
+ "Gender=Masc|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part":87,
414
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Int":95,
415
+ "Gender=Fem|Number=Sing|POS=DET|PronType=Int":90,
416
+ "POS=PRON|PronType=Int":95,
417
+ "Gender=Fem|Number=Plur|POS=DET|PronType=Int":90,
418
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=3|VerbForm=Fin":100,
419
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=3|VerbForm=Fin":100,
420
+ "POS=PART":94,
421
+ "Gender=Fem|Number=Sing|POS=PRON|PronType=Dem":95,
422
+ "Gender=Masc|Number=Sing|POS=DET|PronType=Tot":90,
423
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Dem":95,
424
+ "POS=ADJ":84,
425
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|PronType=Prs":95,
426
+ "Degree=Cmp|Number=Plur|POS=ADJ":84,
427
+ "POS=PUNCT|PunctType=Dash":97,
428
+ "Mood=Sub|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":87,
429
+ "Case=Acc|Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs":95,
430
+ "Mood=Sub|Number=Sing|POS=VERB|Person=3|Tense=Imp|VerbForm=Fin":100,
431
+ "Gender=Fem|Number=Plur|POS=VERB|Tense=Past|VerbForm=Part":100,
432
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
433
+ "Gender=Masc|POS=NOUN":92,
434
+ "Mood=Ind|Number=Sing|POS=VERB|Person=3|Tense=Past|VerbForm=Fin":100,
435
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Int":95,
436
+ "Gender=Masc|NumType=Ord|Number=Plur|POS=ADJ":84,
437
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin":87,
438
+ "POS=PUNCT|PunctType=Colo":97,
439
+ "Gender=Masc|NumType=Card|POS=NUM":93,
440
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
441
+ "Number=Sing|POS=PRON|PronType=Int":95,
442
+ "POS=PUNCT|PunctType=Quot":97,
443
+ "Mood=Imp|Number=Sing|POS=VERB|Person=3|VerbForm=Fin":100,
444
+ "Gender=Fem|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":90,
445
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":90,
446
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin":100,
447
+ "POS=AUX|VerbForm=Ger":87,
448
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|PronType=Prs":95,
449
+ "Mood=Imp|Number=Sing|POS=AUX|Person=3|VerbForm=Fin":87,
450
+ "Number=Plur|POS=PRON|PronType=Ind":95,
451
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Dem":95,
452
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Npr|PronType=Prs":95,
453
+ "Gender=Masc|Number=Plur|POS=PRON|PronType=Int":95,
454
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":87,
455
+ "NumForm=Digit|NumType=Frac|POS=NUM":93,
456
+ "POS=VERB":100,
457
+ "Gender=Fem|Number=Plur|POS=PRON|PronType=Dem":95,
458
+ "Gender=Fem|POS=NOUN":92,
459
+ "Case=Acc,Dat|Number=Sing|POS=PRON|Person=1|PrepCase=Npr|PronType=Prs":95,
460
+ "Mood=Sub|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":100,
461
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Fut|VerbForm=Fin":87,
462
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":87,
463
+ "Mood=Sub|Number=Plur|POS=AUX|Person=3|Tense=Imp|VerbForm=Fin":87,
464
+ "Number=Plur|POS=PRON|Person=1|PronType=Prs":95,
465
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":100,
466
+ "Case=Nom|Number=Sing|POS=PRON|Person=2|Polite=Infm|PronType=Prs":95,
467
+ "POS=X":101,
468
+ "Mood=Cnd|Number=Plur|POS=AUX|Person=1|VerbForm=Fin":87,
469
+ "Number=Sing|POS=DET|PronType=Dem":90,
470
+ "POS=DET":90,
471
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Fut|VerbForm=Fin":100,
472
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":87,
473
+ "POS=DET|PronType=Art":90,
474
+ "Gender=Masc|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":95,
475
+ "NumType=Ord|Number=Sing|POS=ADJ":84,
476
+ "Gender=Fem|Number=Sing|POS=AUX|Tense=Past|VerbForm=Part":87,
477
+ "Number=Plur|Number[psor]=Plur|POS=DET|Person=1|Poss=Yes|PronType=Prs":90,
478
+ "Gender=Fem|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part":87,
479
+ "Gender=Masc|Number=Plur|POS=AUX|Tense=Past|VerbForm=Part":87,
480
+ "Number=Plur|POS=PRON|PronType=Dem":95,
481
+ "Mood=Imp|Number=Plur|POS=VERB|Person=1|VerbForm=Fin":100,
482
+ "POS=PRON|PronType=Ind":95,
483
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":100,
484
+ "Mood=Imp|Number=Plur|POS=VERB|Person=3|VerbForm=Fin":100,
485
+ "Case=Nom|Number=Sing|POS=PRON|Person=1|PronType=Prs":95,
486
+ "Case=Acc|Number=Sing|POS=PRON|Person=1|PrepCase=Pre|PronType=Prs":95,
487
+ "Mood=Ind|Number=Sing|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin":87,
488
+ "Mood=Ind|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":100,
489
+ "POS=PUNCT|PunctSide=Fin|PunctType=Qest":97,
490
+ "NumForm=Digit|NumType=Ord|POS=ADJ":84,
491
+ "Case=Acc|POS=PRON|Person=3|PrepCase=Pre|PronType=Prs|Reflex=Yes":95,
492
+ "NumForm=Digit|NumType=Frac|POS=SYM":99,
493
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":100,
494
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":90,
495
+ "Gender=Masc|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs":95,
496
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Pres|VerbForm=Fin":100,
497
+ "POS=PUNCT|PunctSide=Ini|PunctType=Qest":97,
498
+ "NumType=Card|Number=Sing|POS=NUM":93,
499
+ "Foreign=Yes|POS=PRON|PronType=Int":95,
500
+ "Foreign=Yes|Mood=Ind|POS=VERB|VerbForm=Fin":100,
501
+ "Foreign=Yes|POS=ADP":85,
502
+ "Gender=Masc|Number=Sing|POS=PROPN":96,
503
+ "POS=PUNCT|PunctSide=Ini|PunctType=Excl":97,
504
+ "POS=PUNCT|PunctSide=Fin|PunctType=Excl":97,
505
+ "Mood=Cnd|Number=Sing|POS=AUX|Person=1|VerbForm=Fin":87,
506
+ "Number=Plur|POS=PRON|Person=2|Polite=Form|PronType=Prs":95,
507
+ "Mood=Sub|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":87,
508
+ "POS=PUNCT|PunctSide=Ini|PunctType=Comm":97,
509
+ "POS=PUNCT|PunctSide=Fin|PunctType=Comm":97,
510
+ "Number=Plur|POS=PRON|Person=2|PronType=Prs":95,
511
+ "Mood=Ind|Number=Plur|POS=AUX|Person=2|Tense=Pres|VerbForm=Fin":87,
512
+ "Case=Acc,Dat|Number=Plur|POS=PRON|Person=2|PronType=Prs":95,
513
+ "Mood=Cnd|Number=Sing|POS=VERB|Person=1|VerbForm=Fin":100,
514
+ "Mood=Cnd|Number=Plur|POS=VERB|Person=1|VerbForm=Fin":100,
515
+ "Mood=Ind|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":87,
516
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs":90,
517
+ "Definite=Ind|Gender=Masc|Number=Sing|POS=DET|PronType=Art":90,
518
+ "Number=Sing|POS=PRON|Person=2|Polite=Form|PronType=Prs":95,
519
+ "Gender=Masc|Number=Sing|Number[psor]=Sing|POS=DET|Person=1|Poss=Yes|PronType=Prs":90,
520
+ "Mood=Ind|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":100,
521
+ "POS=VERB|Tense=Past|VerbForm=Part":100,
522
+ "Mood=Imp|Number=Plur|POS=AUX|Person=3|VerbForm=Fin":87,
523
+ "Case=Nom|POS=PRON|Person=3|PronType=Prs":95,
524
+ "Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Past|VerbForm=Fin":87,
525
+ "Gender=Fem|Number=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":95,
526
+ "Gender=Masc|Number=Sing|POS=PRON|PronType=Rel":95,
527
+ "Definite=Ind|Number=Sing|POS=DET|PronType=Art":90,
528
+ "Gender=Masc|Number=Sing|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs":95,
529
+ "Number=Plur|Number[psor]=Plur|POS=PRON|Person=1|Poss=Yes|PronType=Prs":95,
530
+ "POS=AUX|Tense=Past|VerbForm=Part":87,
531
+ "Gender=Fem|NumType=Card|POS=NUM":93,
532
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":87,
533
+ "Mood=Sub|Number=Sing|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":100,
534
+ "Gender=Fem|Number=Plur|POS=PRON|Person=3|Poss=Yes|PronType=Prs":95,
535
+ "Mood=Ind|Number=Sing|POS=AUX|Person=1|Tense=Fut|VerbForm=Fin":87,
536
+ "Mood=Ind|Number=Plur|POS=AUX|Person=3|Tense=Past|VerbForm=Fin":87,
537
+ "AdvType=Tim|Degree=Cmp|POS=ADV":86,
538
+ "Case=Acc|Number=Sing|POS=PRON|Person=2|Polite=Infm|PrepCase=Pre|PronType=Prs":95,
539
+ "POS=DET|PronType=Rel":90,
540
+ "Definite=Ind|Gender=Fem|Number=Plur|POS=DET|PronType=Art":90,
541
+ "Mood=Ind|Number=Plur|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin":100,
542
+ "POS=INTJ":91,
543
+ "Mood=Sub|Number=Sing|POS=AUX|Person=1|Tense=Pres|VerbForm=Fin":87,
544
+ "POS=VERB|VerbForm=Fin":100,
545
+ "Mood=Ind|Number=Plur|POS=VERB|Person=3|Tense=Past|VerbForm=Fin":100,
546
+ "Definite=Ind|Gender=Fem|Number=Sing|POS=DET|PronType=Art":90,
547
+ "Mood=Sub|Number=Plur|POS=AUX|Person=1|Tense=Imp|VerbForm=Fin":87,
548
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=PRON|Person=3|Poss=Yes|PronType=Prs":95,
549
+ "Mood=Sub|Number=Sing|POS=VERB|Person=2|Tense=Pres|VerbForm=Fin":100,
550
+ "Case=Acc|POS=PRON|Person=3|PronType=Prs|Reflex=Yes":95,
551
+ "Foreign=Yes|POS=NOUN":92,
552
+ "Foreign=Yes|Mood=Ind|Number=Sing|POS=AUX|Person=3|Tense=Pres|VerbForm=Fin":87,
553
+ "Foreign=Yes|Gender=Masc|Number=Sing|POS=PRON|Person=3|PronType=Prs":95,
554
+ "Foreign=Yes|POS=SCONJ":98,
555
+ "Foreign=Yes|Gender=Fem|Number=Sing|POS=DET|PronType=Art":90,
556
+ "Gender=Masc|POS=SYM":99,
557
+ "Gender=Fem|Number=Sing|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":90,
558
+ "Number=Sing|POS=DET|Person=3|Poss=Yes|PronType=Prs":90,
559
+ "Gender=Masc|Number=Plur|Number[psor]=Sing|POS=DET|Person=2|Poss=Yes|PronType=Prs":90,
560
+ "Gender=Fem|Number=Sing|POS=PROPN":96,
561
+ "Mood=Sub|Number=Plur|POS=VERB|Person=1|Tense=Imp|VerbForm=Fin":100,
562
+ "Definite=Def|Foreign=Yes|Gender=Masc|Number=Sing|POS=DET|PronType=Art":90,
563
+ "Foreign=Yes|POS=VERB":100,
564
+ "Foreign=Yes|POS=ADJ":84,
565
+ "Foreign=Yes|POS=DET":90,
566
+ "Foreign=Yes|POS=ADV":86,
567
+ "POS=PUNCT|PunctSide=Fin|Punta d'aignctType=Brck":97,
568
+ "Degree=Cmp|POS=ADJ":84,
569
+ "AdvType=Tim|POS=SYM":99,
570
+ "Number=Plur|POS=DET|PronType=Dem":90,
571
+ "Mood=Ind|Number=Sing|POS=VERB|Person=2|Tense=Fut|VerbForm=Fin":100
572
+ },
573
+ "overwrite":true
574
+ }
morphologizer/model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:584bfa3fc740d6b0124dc0bade8e00fe19d003694df413d15c71a2256e5e771a
3
+ size 7070904
ner/cfg ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "moves":null,
3
+ "update_with_oracle_cut_size":100,
4
+ "multitasks":[
5
+
6
+ ],
7
+ "min_action_freq":1,
8
+ "learn_tokens":false,
9
+ "beam_width":1,
10
+ "beam_density":0.0,
11
+ "beam_update_prob":0.0,
12
+ "incorrect_spans_key":null
13
+ }
ner/model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b37473fafc2392bf6f98d02fde498925b6c3151b7d3e1d5d5196e2e0a939b31
3
+ size 7098032
ner/moves ADDED
@@ -0,0 +1 @@
 
 
1
+ ��moves� {"0":{},"1":{"ORG":17258,"PER":11788,"LOC":10825,"MISC":8394,"ZIP":-1,"EMAIL":-2,"TELEPHONE":-3,"VEHICLE":-4,"ID":-5,"FINANCIAL":-6},"2":{"ORG":17258,"PER":11788,"LOC":10825,"MISC":8394,"ZIP":-1,"EMAIL":-2,"TELEPHONE":-3,"VEHICLE":-4,"ID":-5,"FINANCIAL":-6},"3":{"ORG":17258,"PER":11788,"LOC":10825,"MISC":8394,"ZIP":-1,"EMAIL":-2,"TELEPHONE":-3,"VEHICLE":-4,"ID":-5,"FINANCIAL":-6},"4":{"ORG":17258,"PER":11788,"LOC":10825,"MISC":8394,"":1,"ZIP":-1,"EMAIL":-2,"TELEPHONE":-3,"VEHICLE":-4,"ID":-5,"FINANCIAL":-6},"5":{"":1}}�cfg��neg_key�
parser/cfg ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "moves":null,
3
+ "update_with_oracle_cut_size":100,
4
+ "multitasks":[
5
+
6
+ ],
7
+ "min_action_freq":30,
8
+ "learn_tokens":false,
9
+ "beam_width":1,
10
+ "beam_density":0.0,
11
+ "beam_update_prob":0.0,
12
+ "incorrect_spans_key":null
13
+ }
parser/model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7010f9c08dd59e4c1147056d3e57bf07f9ec34f860ed3b83586bdd93869e23a7
3
+ size 7269438
parser/moves ADDED
@@ -0,0 +1 @@
 
 
1
+ ��moves�\{"0":{"":258157},"1":{"":195402},"2":{"det":72384,"case":65325,"punct":19483,"nsubj":19448,"aux":14225,"mark":12801,"cc":12635,"obj":9225,"advmod":8826,"obl":5440,"amod":4650,"cop":3861,"nummod":3859,"nmod":2488,"advcl":870,"iobj":626,"expl:pass":428,"ccomp":423,"fixed||case":265,"parataxis":228,"csubj":226,"xcomp":129,"det||case":38,"case||case":38,"advcl||xcomp":30,"dep":0},"3":{"nmod":35128,"punct":26112,"obj":22359,"amod":16770,"flat":15296,"conj":14234,"obl":13100,"acl":8144,"fixed":7303,"appos":6631,"advcl":4904,"advmod":4252,"nsubj":3999,"ccomp":3299,"xcomp":2939,"compound":2931,"case":1858,"det":1677,"nummod":1053,"csubj":712,"iobj":668,"aux":325,"parataxis":254,"flat||det":229,"mark":217,"cop":159,"nmod||det":135,"appos||det":90,"dep":59,"fixed||case":58,"conj||xcomp":55,"obj||xcomp":50,"advcl||xcomp":46,"fixed||cc":37},"4":{"ROOT":14319}}�cfg��neg_key�
tok2vec/cfg ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+
3
+ }
tok2vec/model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5802628dc78a4411832a4f9740a3a81fa944fac7eadce24ff9dceceb707b3273
3
+ size 6960804
tokenizer ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ ��prefix_search� �^§|^%|^=|^—|^–|^-|^\+(?![0-9])|^…|^……|^,|^:|^;|^\!|^\?|^¿|^؟|^¡|^\(|^\)|^\[|^\]|^\{|^\}|^<|^>|^_|^#|^\*|^&|^。|^?|^!|^,|^、|^;|^:|^~|^·|^।|^،|^۔|^؛|^٪|^\.\.+|^…|^\'|^"|^”|^“|^`|^‘|^´|^’|^‚|^,|^„|^»|^«|^「|^」|^『|^』|^(|^)|^〔|^〕|^【|^】|^《|^》|^〈|^〉|^\$|^£|^€|^¥|^฿|^US\$|^C\$|^A\$|^₽|^﷼|^₴|^₠|^₡|^₢|^₣|^₤|^₥|^₦|^₧|^₨|^₩|^₪|^₫|^€|^₭|^₮|^₯|^₰|^₱|^₲|^₳|^₴|^₵|^₶|^₷|^₸|^₹|^₺|^₻|^₼|^₽|^₾|^₿|^[\u00A6\u00A9\u00AE\u00B0\u0482\u058D\u058E\u060E\u060F\u06DE\u06E9\u06FD\u06FE\u07F6\u09FA\u0B70\u0BF3-\u0BF8\u0BFA\u0C7F\u0D4F\u0D79\u0F01-\u0F03\u0F13\u0F15-\u0F17\u0F1A-\u0F1F\u0F34\u0F36\u0F38\u0FBE-\u0FC5\u0FC7-\u0FCC\u0FCE\u0FCF\u0FD5-\u0FD8\u109E\u109F\u1390-\u1399\u1940\u19DE-\u19FF\u1B61-\u1B6A\u1B74-\u1B7C\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116\u2117\u211E-\u2123\u2125\u2127\u2129\u212E\u213A\u213B\u214A\u214C\u214D\u214F\u218A\u218B\u2195-\u2199\u219C-\u219F\u21A1\u21A2\u21A4\u21A5\u21A7-\u21AD\u21AF-\u21CD\u21D0\u21D1\u21D3\u21D5-\u21F3\u2300-\u2307\u230C-\u231F\u2322-\u2328\u232B-\u237B\u237D-\u239A\u23B4-\u23DB\u23E2-\u2426\u2440-\u244A\u249C-\u24E9\u2500-\u25B6\u25B8-\u25C0\u25C2-\u25F7\u2600-\u266E\u2670-\u2767\u2794-\u27BF\u2800-\u28FF\u2B00-\u2B2F\u2B45\u2B46\u2B4D-\u2B73\u2B76-\u2B95\u2B98-\u2BC8\u2BCA-\u2BFE\u2CE5-\u2CEA\u2E80-\u2E99\u2E9B-\u2EF3\u2F00-\u2FD5\u2FF0-\u2FFB\u3004\u3012\u3013\u3020\u3036\u3037\u303E\u303F\u3190\u3191\u3196-\u319F\u31C0-\u31E3\u3200-\u321E\u322A-\u3247\u3250\u3260-\u327F\u328A-\u32B0\u32C0-\u32FE\u3300-\u33FF\u4DC0-\u4DFF\uA490-\uA4C6\uA828-\uA82B\uA836\uA837\uA839\uAA77-\uAA79\uFDFD\uFFE4\uFFE8\uFFED\uFFEE\uFFFC\uFFFD\U00010137-\U0001013F\U00010179-\U00010189\U0001018C-\U0001018E\U00010190-\U0001019B\U000101A0\U000101D0-\U000101FC\U00010877\U00010878\U00010AC8\U0001173F\U00016B3C-\U00016B3F\U00016B45\U0001BC9C\U0001D000-\U0001D0F5\U0001D100-\U0001D126\U0001D129-\U0001D164\U0001D16A-\U0001D16C\U0001D183\U0001D184\U0001D18C-\U0001D1A9\U0001D1AE-\U0001D1E8\U0001D200-\U0001D241\U0001D245\U0001D300-\U0001D356\U0001D800-\U0001D9FF\U0001DA37-\U0001DA3A\U0001DA6D-\U0001DA74\U0001DA76-\U0001DA83\U0001DA85\U0001DA86\U0001ECAC\U0001F000-\U0001F02B\U0001F030-\U0001F093\U0001F0A0-\U0001F0AE\U0001F0B1-\U0001F0BF\U0001F0C1-\U0001F0CF\U0001F0D1-\U0001F0F5\U0001F110-\U0001F16B\U0001F170-\U0001F1AC\U0001F1E6-\U0001F202\U0001F210-\U0001F23B\U0001F240-\U0001F248\U0001F250\U0001F251\U0001F260-\U0001F265\U0001F300-\U0001F3FA\U0001F400-\U0001F6D4\U0001F6E0-\U0001F6EC\U0001F6F0-\U0001F6F9\U0001F700-\U0001F773\U0001F780-\U0001F7D8\U0001F800-\U0001F80B\U0001F810-\U0001F847\U0001F850-\U0001F859\U0001F860-\U0001F887\U0001F890-\U0001F8AD\U0001F900-\U0001F90B\U0001F910-\U0001F93E\U0001F940-\U0001F970\U0001F973-\U0001F976\U0001F97A\U0001F97C-\U0001F9A2\U0001F9B0-\U0001F9B9\U0001F9C0-\U0001F9C2\U0001F9D0-\U0001F9FF\U0001FA60-\U0001FA6D]�suffix_search�2…$|……$|,$|:$|;$|\!$|\?$|¿$|؟$|¡$|\($|\)$|\[$|\]$|\{$|\}$|<$|>$|_$|#$|\*$|&$|。$|?$|!$|,$|、$|;$|:$|~$|·$|।$|،$|۔$|؛$|٪$|\.\.+$|…$|\'$|"$|”$|“$|`$|‘$|´$|’$|‚$|,$|„$|»$|«$|「$|」$|『$|』$|($|)$|〔$|〕$|【$|】$|《$|》$|〈$|〉$|[\u00A6\u00A9\u00AE\u00B0\u0482\u058D\u058E\u060E\u060F\u06DE\u06E9\u06FD\u06FE\u07F6\u09FA\u0B70\u0BF3-\u0BF8\u0BFA\u0C7F\u0D4F\u0D79\u0F01-\u0F03\u0F13\u0F15-\u0F17\u0F1A-\u0F1F\u0F34\u0F36\u0F38\u0FBE-\u0FC5\u0FC7-\u0FCC\u0FCE\u0FCF\u0FD5-\u0FD8\u109E\u109F\u1390-\u1399\u1940\u19DE-\u19FF\u1B61-\u1B6A\u1B74-\u1B7C\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116\u2117\u211E-\u2123\u2125\u2127\u2129\u212E\u213A\u213B\u214A\u214C\u214D\u214F\u218A\u218B\u2195-\u2199\u219C-\u219F\u21A1\u21A2\u21A4\u21A5\u21A7-\u21AD\u21AF-\u21CD\u21D0\u21D1\u21D3\u21D5-\u21F3\u2300-\u2307\u230C-\u231F\u2322-\u2328\u232B-\u237B\u237D-\u239A\u23B4-\u23DB\u23E2-\u2426\u2440-\u244A\u249C-\u24E9\u2500-\u25B6\u25B8-\u25C0\u25C2-\u25F7\u2600-\u266E\u2670-\u2767\u2794-\u27BF\u2800-\u28FF\u2B00-\u2B2F\u2B45\u2B46\u2B4D-\u2B73\u2B76-\u2B95\u2B98-\u2BC8\u2BCA-\u2BFE\u2CE5-\u2CEA\u2E80-\u2E99\u2E9B-\u2EF3\u2F00-\u2FD5\u2FF0-\u2FFB\u3004\u3012\u3013\u3020\u3036\u3037\u303E\u303F\u3190\u3191\u3196-\u319F\u31C0-\u31E3\u3200-\u321E\u322A-\u3247\u3250\u3260-\u327F\u328A-\u32B0\u32C0-\u32FE\u3300-\u33FF\u4DC0-\u4DFF\uA490-\uA4C6\uA828-\uA82B\uA836\uA837\uA839\uAA77-\uAA79\uFDFD\uFFE4\uFFE8\uFFED\uFFEE\uFFFC\uFFFD\U00010137-\U0001013F\U00010179-\U00010189\U0001018C-\U0001018E\U00010190-\U0001019B\U000101A0\U000101D0-\U000101FC\U00010877\U00010878\U00010AC8\U0001173F\U00016B3C-\U00016B3F\U00016B45\U0001BC9C\U0001D000-\U0001D0F5\U0001D100-\U0001D126\U0001D129-\U0001D164\U0001D16A-\U0001D16C\U0001D183\U0001D184\U0001D18C-\U0001D1A9\U0001D1AE-\U0001D1E8\U0001D200-\U0001D241\U0001D245\U0001D300-\U0001D356\U0001D800-\U0001D9FF\U0001DA37-\U0001DA3A\U0001DA6D-\U0001DA74\U0001DA76-\U0001DA83\U0001DA85\U0001DA86\U0001ECAC\U0001F000-\U0001F02B\U0001F030-\U0001F093\U0001F0A0-\U0001F0AE\U0001F0B1-\U0001F0BF\U0001F0C1-\U0001F0CF\U0001F0D1-\U0001F0F5\U0001F110-\U0001F16B\U0001F170-\U0001F1AC\U0001F1E6-\U0001F202\U0001F210-\U0001F23B\U0001F240-\U0001F248\U0001F250\U0001F251\U0001F260-\U0001F265\U0001F300-\U0001F3FA\U0001F400-\U0001F6D4\U0001F6E0-\U0001F6EC\U0001F6F0-\U0001F6F9\U0001F700-\U0001F773\U0001F780-\U0001F7D8\U0001F800-\U0001F80B\U0001F810-\U0001F847\U0001F850-\U0001F859\U0001F860-\U0001F887\U0001F890-\U0001F8AD\U0001F900-\U0001F90B\U0001F910-\U0001F93E\U0001F940-\U0001F970\U0001F973-\U0001F976\U0001F97A\U0001F97C-\U0001F9A2\U0001F9B0-\U0001F9B9\U0001F9C0-\U0001F9C2\U0001F9D0-\U0001F9FF\U0001FA60-\U0001FA6D]$|-$|—$|–$|(?<=[0-9])\+$|(?<=°[FfCcKk])\.$|(?<=[0-9])(?:\$|£|€|¥|฿|US\$|C\$|A\$|₽|﷼|₴|₠|₡|₢|₣|₤|₥|₦|₧|₨|₩|₪|₫|€|₭|₮|₯|₰|₱|₲|₳|₴|₵|₶|₷|₸|₹|₺|₻|₼|₽|₾|₿)$|(?<=[0-9])(?:km|km²|km³|m|m²|m³|dm|dm²|dm³|cm|cm²|cm³|mm|mm²|mm³|ha|µm|nm|yd|in|ft|kg|g|mg|µg|t|lb|oz|m/s|km/h|kmh|mph|hPa|Pa|mbar|mb|MB|kb|KB|gb|GB|tb|TB|T|G|M|K|км|км²|км³|м|м²|м³|дм|дм²|дм³|см|см²|см³|мм|мм²|мм³|нм|кг|г|мг|м/с|км/ч|кПа|Па|мбар|Кб|КБ|кб|Мб|МБ|мб|Гб|ГБ|гб|Тб|ТБ|тбكم|كم²|كم³|م|م²|م³|سم|سم²|سم³|مم|مم²|مم³|كم|غرام|جرام|جم|كغ|ملغ|كوب|اكواب)$|(?<=[0-9a-z\uFF41-\uFF5A\u00DF-\u00F6\u00F8-\u00FF\u0101\u0103\u0105\u0107\u0109\u010B\u010D\u010F\u0111\u0113\u0115\u0117\u0119\u011B\u011D\u011F\u0121\u0123\u0125\u0127\u0129\u012B\u012D\u012F\u0131\u0133\u0135\u0137\u0138\u013A\u013C\u013E\u0140\u0142\u0144\u0146\u0148\u0149\u014B\u014D\u014F\u0151\u0153\u0155\u0157\u0159\u015B\u015D\u015F\u0161\u0163\u0165\u0167\u0169\u016B\u016D\u016F\u0171\u0173\u0175\u0177\u017A\u017C\u017E\u017F\u0180\u0183\u0185\u0188\u018C\u018D\u0192\u0195\u0199-\u019B\u019E\u01A1\u01A3\u01A5\u01A8\u01AA\u01AB\u01AD\u01B0\u01B4\u01B6\u01B9\u01BA\u01BD-\u01BF\u01C6\u01C9\u01CC\u01CE\u01D0\u01D2\u01D4\u01D6\u01D8\u01DA\u01DC\u01DD\u01DF\u01E1\u01E3\u01E5\u01E7\u01E9\u01EB\u01ED\u01EF\u01F0\u01F3\u01F5\u01F9\u01FB\u01FD\u01FF\u0201\u0203\u0205\u0207\u0209\u020B\u020D\u020F\u0211\u0213\u0215\u0217\u0219\u021B\u021D\u021F\u0221\u0223\u0225\u0227\u0229\u022B\u022D\u022F\u0231\u0233-\u0239\u023C\u023F\u0240\u0242\u0247\u0249\u024B\u024D\u024F\u2C61\u2C65\u2C66\u2C68\u2C6A\u2C6C\u2C71\u2C73\u2C74\u2C76-\u2C7B\uA723\uA725\uA727\uA729\uA72B\uA72D\uA72F-\uA731\uA733\uA735\uA737\uA739\uA73B\uA73D\uA73F\uA741\uA743\uA745\uA747\uA749\uA74B\uA74D\uA74F\uA751\uA753\uA755\uA757\uA759\uA75B\uA75D\uA75F\uA761\uA763\uA765\uA767\uA769\uA76B\uA76D\uA76F\uA771-\uA778\uA77A\uA77C\uA77F\uA781\uA783\uA785\uA787\uA78C\uA78E\uA791\uA793-\uA795\uA797\uA799\uA79B\uA79D\uA79F\uA7A1\uA7A3\uA7A5\uA7A7\uA7A9\uA7AF\uA7B5\uA7B7\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E01\u1E03\u1E05\u1E07\u1E09\u1E0B\u1E0D\u1E0F\u1E11\u1E13\u1E15\u1E17\u1E19\u1E1B\u1E1D\u1E1F\u1E21\u1E23\u1E25\u1E27\u1E29\u1E2B\u1E2D\u1E2F\u1E31\u1E33\u1E35\u1E37\u1E39\u1E3B\u1E3D\u1E3F\u1E41\u1E43\u1E45\u1E47\u1E49\u1E4B\u1E4D\u1E4F\u1E51\u1E53\u1E55\u1E57\u1E59\u1E5B\u1E5D\u1E5F\u1E61\u1E63\u1E65\u1E67\u1E69\u1E6B\u1E6D\u1E6F\u1E71\u1E73\u1E75\u1E77\u1E79\u1E7B\u1E7D\u1E7F\u1E81\u1E83\u1E85\u1E87\u1E89\u1E8B\u1E8D\u1E8F\u1E91\u1E93\u1E95-\u1E9D\u1E9F\u1EA1\u1EA3\u1EA5\u1EA7\u1EA9\u1EAB\u1EAD\u1EAF\u1EB1\u1EB3\u1EB5\u1EB7\u1EB9\u1EBB\u1EBD\u1EBF\u1EC1\u1EC3\u1EC5\u1EC7\u1EC9\u1ECB\u1ECD\u1ECF\u1ED1\u1ED3\u1ED5\u1ED7\u1ED9\u1EDB\u1EDD\u1EDF\u1EE1\u1EE3\u1EE5\u1EE7\u1EE9\u1EEB\u1EED\u1EEF\u1EF1\u1EF3\u1EF5\u1EF7\u1EF9\u1EFB\u1EFD\u1EFFёа-яәөүҗңһα-ωάέίόώήύа-щюяіїєґѓѕјљњќѐѝ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F%²\-\+…|……|,|:|;|\!|\?|¿|؟|¡|\(|\)|\[|\]|\{|\}|<|>|_|#|\*|&|。|?|!|,|、|;|:|~|·|।|،|۔|؛|٪(?:\'"”“`‘´’‚,„»«「」『』()〔〕【】《》〈〉)])\.$|(?<=[A-Z\uFF21-\uFF3A\u00C0-\u00D6\u00D8-\u00DE\u0100\u0102\u0104\u0106\u0108\u010A\u010C\u010E\u0110\u0112\u0114\u0116\u0118\u011A\u011C\u011E\u0120\u0122\u0124\u0126\u0128\u012A\u012C\u012E\u0130\u0132\u0134\u0136\u0139\u013B\u013D\u013F\u0141\u0143\u0145\u0147\u014A\u014C\u014E\u0150\u0152\u0154\u0156\u0158\u015A\u015C\u015E\u0160\u0162\u0164\u0166\u0168\u016A\u016C\u016E\u0170\u0172\u0174\u0176\u0178\u0179\u017B\u017D\u0181\u0182\u0184\u0186\u0187\u0189-\u018B\u018E-\u0191\u0193\u0194\u0196-\u0198\u019C\u019D\u019F\u01A0\u01A2\u01A4\u01A6\u01A7\u01A9\u01AC\u01AE\u01AF\u01B1-\u01B3\u01B5\u01B7\u01B8\u01BC\u01C4\u01C7\u01CA\u01CD\u01CF\u01D1\u01D3\u01D5\u01D7\u01D9\u01DB\u01DE\u01E0\u01E2\u01E4\u01E6\u01E8\u01EA\u01EC\u01EE\u01F1\u01F4\u01F6-\u01F8\u01FA\u01FC\u01FE\u0200\u0202\u0204\u0206\u0208\u020A\u020C\u020E\u0210\u0212\u0214\u0216\u0218\u021A\u021C\u021E\u0220\u0222\u0224\u0226\u0228\u022A\u022C\u022E\u0230\u0232\u023A\u023B\u023D\u023E\u0241\u0243-\u0246\u0248\u024A\u024C\u024E\u2C60\u2C62-\u2C64\u2C67\u2C69\u2C6B\u2C6D-\u2C70\u2C72\u2C75\u2C7E\u2C7F\uA722\uA724\uA726\uA728\uA72A\uA72C\uA72E\uA732\uA734\uA736\uA738\uA73A\uA73C\uA73E\uA740\uA742\uA744\uA746\uA748\uA74A\uA74C\uA74E\uA750\uA752\uA754\uA756\uA758\uA75A\uA75C\uA75E\uA760\uA762\uA764\uA766\uA768\uA76A\uA76C\uA76E\uA779\uA77B\uA77D\uA77E\uA780\uA782\uA784\uA786\uA78B\uA78D\uA790\uA792\uA796\uA798\uA79A\uA79C\uA79E\uA7A0\uA7A2\uA7A4\uA7A6\uA7A8\uA7AA-\uA7AE\uA7B0-\uA7B4\uA7B6\uA7B8\u1E00\u1E02\u1E04\u1E06\u1E08\u1E0A\u1E0C\u1E0E\u1E10\u1E12\u1E14\u1E16\u1E18\u1E1A\u1E1C\u1E1E\u1E20\u1E22\u1E24\u1E26\u1E28\u1E2A\u1E2C\u1E2E\u1E30\u1E32\u1E34\u1E36\u1E38\u1E3A\u1E3C\u1E3E\u1E40\u1E42\u1E44\u1E46\u1E48\u1E4A\u1E4C\u1E4E\u1E50\u1E52\u1E54\u1E56\u1E58\u1E5A\u1E5C\u1E5E\u1E60\u1E62\u1E64\u1E66\u1E68\u1E6A\u1E6C\u1E6E\u1E70\u1E72\u1E74\u1E76\u1E78\u1E7A\u1E7C\u1E7E\u1E80\u1E82\u1E84\u1E86\u1E88\u1E8A\u1E8C\u1E8E\u1E90\u1E92\u1E94\u1E9E\u1EA0\u1EA2\u1EA4\u1EA6\u1EA8\u1EAA\u1EAC\u1EAE\u1EB0\u1EB2\u1EB4\u1EB6\u1EB8\u1EBA\u1EBC\u1EBE\u1EC0\u1EC2\u1EC4\u1EC6\u1EC8\u1ECA\u1ECC\u1ECE\u1ED0\u1ED2\u1ED4\u1ED6\u1ED8\u1EDA\u1EDC\u1EDE\u1EE0\u1EE2\u1EE4\u1EE6\u1EE8\u1EEA\u1EEC\u1EEE\u1EF0\u1EF2\u1EF4\u1EF6\u1EF8\u1EFA\u1EFC\u1EFEЁА-ЯӘӨҮҖҢҺΑ-ΩΆΈΊΌΏΉΎА-ЩЮЯІЇЄҐЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F][A-Z\uFF21-\uFF3A\u00C0-\u00D6\u00D8-\u00DE\u0100\u0102\u0104\u0106\u0108\u010A\u010C\u010E\u0110\u0112\u0114\u0116\u0118\u011A\u011C\u011E\u0120\u0122\u0124\u0126\u0128\u012A\u012C\u012E\u0130\u0132\u0134\u0136\u0139\u013B\u013D\u013F\u0141\u0143\u0145\u0147\u014A\u014C\u014E\u0150\u0152\u0154\u0156\u0158\u015A\u015C\u015E\u0160\u0162\u0164\u0166\u0168\u016A\u016C\u016E\u0170\u0172\u0174\u0176\u0178\u0179\u017B\u017D\u0181\u0182\u0184\u0186\u0187\u0189-\u018B\u018E-\u0191\u0193\u0194\u0196-\u0198\u019C\u019D\u019F\u01A0\u01A2\u01A4\u01A6\u01A7\u01A9\u01AC\u01AE\u01AF\u01B1-\u01B3\u01B5\u01B7\u01B8\u01BC\u01C4\u01C7\u01CA\u01CD\u01CF\u01D1\u01D3\u01D5\u01D7\u01D9\u01DB\u01DE\u01E0\u01E2\u01E4\u01E6\u01E8\u01EA\u01EC\u01EE\u01F1\u01F4\u01F6-\u01F8\u01FA\u01FC\u01FE\u0200\u0202\u0204\u0206\u0208\u020A\u020C\u020E\u0210\u0212\u0214\u0216\u0218\u021A\u021C\u021E\u0220\u0222\u0224\u0226\u0228\u022A\u022C\u022E\u0230\u0232\u023A\u023B\u023D\u023E\u0241\u0243-\u0246\u0248\u024A\u024C\u024E\u2C60\u2C62-\u2C64\u2C67\u2C69\u2C6B\u2C6D-\u2C70\u2C72\u2C75\u2C7E\u2C7F\uA722\uA724\uA726\uA728\uA72A\uA72C\uA72E\uA732\uA734\uA736\uA738\uA73A\uA73C\uA73E\uA740\uA742\uA744\uA746\uA748\uA74A\uA74C\uA74E\uA750\uA752\uA754\uA756\uA758\uA75A\uA75C\uA75E\uA760\uA762\uA764\uA766\uA768\uA76A\uA76C\uA76E\uA779\uA77B\uA77D\uA77E\uA780\uA782\uA784\uA786\uA78B\uA78D\uA790\uA792\uA796\uA798\uA79A\uA79C\uA79E\uA7A0\uA7A2\uA7A4\uA7A6\uA7A8\uA7AA-\uA7AE\uA7B0-\uA7B4\uA7B6\uA7B8\u1E00\u1E02\u1E04\u1E06\u1E08\u1E0A\u1E0C\u1E0E\u1E10\u1E12\u1E14\u1E16\u1E18\u1E1A\u1E1C\u1E1E\u1E20\u1E22\u1E24\u1E26\u1E28\u1E2A\u1E2C\u1E2E\u1E30\u1E32\u1E34\u1E36\u1E38\u1E3A\u1E3C\u1E3E\u1E40\u1E42\u1E44\u1E46\u1E48\u1E4A\u1E4C\u1E4E\u1E50\u1E52\u1E54\u1E56\u1E58\u1E5A\u1E5C\u1E5E\u1E60\u1E62\u1E64\u1E66\u1E68\u1E6A\u1E6C\u1E6E\u1E70\u1E72\u1E74\u1E76\u1E78\u1E7A\u1E7C\u1E7E\u1E80\u1E82\u1E84\u1E86\u1E88\u1E8A\u1E8C\u1E8E\u1E90\u1E92\u1E94\u1E9E\u1EA0\u1EA2\u1EA4\u1EA6\u1EA8\u1EAA\u1EAC\u1EAE\u1EB0\u1EB2\u1EB4\u1EB6\u1EB8\u1EBA\u1EBC\u1EBE\u1EC0\u1EC2\u1EC4\u1EC6\u1EC8\u1ECA\u1ECC\u1ECE\u1ED0\u1ED2\u1ED4\u1ED6\u1ED8\u1EDA\u1EDC\u1EDE\u1EE0\u1EE2\u1EE4\u1EE6\u1EE8\u1EEA\u1EEC\u1EEE\u1EF0\u1EF2\u1EF4\u1EF6\u1EF8\u1EFA\u1EFC\u1EFEЁА-ЯӘӨҮҖҢҺΑ-ΩΆΈΊΌΏΉΎА-ЩЮЯІЇЄҐЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F])\.$�infix_finditer�>Q\.\.+|…|[\u00A6\u00A9\u00AE\u00B0\u0482\u058D\u058E\u060E\u060F\u06DE\u06E9\u06FD\u06FE\u07F6\u09FA\u0B70\u0BF3-\u0BF8\u0BFA\u0C7F\u0D4F\u0D79\u0F01-\u0F03\u0F13\u0F15-\u0F17\u0F1A-\u0F1F\u0F34\u0F36\u0F38\u0FBE-\u0FC5\u0FC7-\u0FCC\u0FCE\u0FCF\u0FD5-\u0FD8\u109E\u109F\u1390-\u1399\u1940\u19DE-\u19FF\u1B61-\u1B6A\u1B74-\u1B7C\u2100\u2101\u2103-\u2106\u2108\u2109\u2114\u2116\u2117\u211E-\u2123\u2125\u2127\u2129\u212E\u213A\u213B\u214A\u214C\u214D\u214F\u218A\u218B\u2195-\u2199\u219C-\u219F\u21A1\u21A2\u21A4\u21A5\u21A7-\u21AD\u21AF-\u21CD\u21D0\u21D1\u21D3\u21D5-\u21F3\u2300-\u2307\u230C-\u231F\u2322-\u2328\u232B-\u237B\u237D-\u239A\u23B4-\u23DB\u23E2-\u2426\u2440-\u244A\u249C-\u24E9\u2500-\u25B6\u25B8-\u25C0\u25C2-\u25F7\u2600-\u266E\u2670-\u2767\u2794-\u27BF\u2800-\u28FF\u2B00-\u2B2F\u2B45\u2B46\u2B4D-\u2B73\u2B76-\u2B95\u2B98-\u2BC8\u2BCA-\u2BFE\u2CE5-\u2CEA\u2E80-\u2E99\u2E9B-\u2EF3\u2F00-\u2FD5\u2FF0-\u2FFB\u3004\u3012\u3013\u3020\u3036\u3037\u303E\u303F\u3190\u3191\u3196-\u319F\u31C0-\u31E3\u3200-\u321E\u322A-\u3247\u3250\u3260-\u327F\u328A-\u32B0\u32C0-\u32FE\u3300-\u33FF\u4DC0-\u4DFF\uA490-\uA4C6\uA828-\uA82B\uA836\uA837\uA839\uAA77-\uAA79\uFDFD\uFFE4\uFFE8\uFFED\uFFEE\uFFFC\uFFFD\U00010137-\U0001013F\U00010179-\U00010189\U0001018C-\U0001018E\U00010190-\U0001019B\U000101A0\U000101D0-\U000101FC\U00010877\U00010878\U00010AC8\U0001173F\U00016B3C-\U00016B3F\U00016B45\U0001BC9C\U0001D000-\U0001D0F5\U0001D100-\U0001D126\U0001D129-\U0001D164\U0001D16A-\U0001D16C\U0001D183\U0001D184\U0001D18C-\U0001D1A9\U0001D1AE-\U0001D1E8\U0001D200-\U0001D241\U0001D245\U0001D300-\U0001D356\U0001D800-\U0001D9FF\U0001DA37-\U0001DA3A\U0001DA6D-\U0001DA74\U0001DA76-\U0001DA83\U0001DA85\U0001DA86\U0001ECAC\U0001F000-\U0001F02B\U0001F030-\U0001F093\U0001F0A0-\U0001F0AE\U0001F0B1-\U0001F0BF\U0001F0C1-\U0001F0CF\U0001F0D1-\U0001F0F5\U0001F110-\U0001F16B\U0001F170-\U0001F1AC\U0001F1E6-\U0001F202\U0001F210-\U0001F23B\U0001F240-\U0001F248\U0001F250\U0001F251\U0001F260-\U0001F265\U0001F300-\U0001F3FA\U0001F400-\U0001F6D4\U0001F6E0-\U0001F6EC\U0001F6F0-\U0001F6F9\U0001F700-\U0001F773\U0001F780-\U0001F7D8\U0001F800-\U0001F80B\U0001F810-\U0001F847\U0001F850-\U0001F859\U0001F860-\U0001F887\U0001F890-\U0001F8AD\U0001F900-\U0001F90B\U0001F910-\U0001F93E\U0001F940-\U0001F970\U0001F973-\U0001F976\U0001F97A\U0001F97C-\U0001F9A2\U0001F9B0-\U0001F9B9\U0001F9C0-\U0001F9C2\U0001F9D0-\U0001F9FF\U0001FA60-\U0001FA6D]|(?<=[0-9])[+\-\*^](?=[0-9-])|(?<=[a-z\uFF41-\uFF5A\u00DF-\u00F6\u00F8-\u00FF\u0101\u0103\u0105\u0107\u0109\u010B\u010D\u010F\u0111\u0113\u0115\u0117\u0119\u011B\u011D\u011F\u0121\u0123\u0125\u0127\u0129\u012B\u012D\u012F\u0131\u0133\u0135\u0137\u0138\u013A\u013C\u013E\u0140\u0142\u0144\u0146\u0148\u0149\u014B\u014D\u014F\u0151\u0153\u0155\u0157\u0159\u015B\u015D\u015F\u0161\u0163\u0165\u0167\u0169\u016B\u016D\u016F\u0171\u0173\u0175\u0177\u017A\u017C\u017E\u017F\u0180\u0183\u0185\u0188\u018C\u018D\u0192\u0195\u0199-\u019B\u019E\u01A1\u01A3\u01A5\u01A8\u01AA\u01AB\u01AD\u01B0\u01B4\u01B6\u01B9\u01BA\u01BD-\u01BF\u01C6\u01C9\u01CC\u01CE\u01D0\u01D2\u01D4\u01D6\u01D8\u01DA\u01DC\u01DD\u01DF\u01E1\u01E3\u01E5\u01E7\u01E9\u01EB\u01ED\u01EF\u01F0\u01F3\u01F5\u01F9\u01FB\u01FD\u01FF\u0201\u0203\u0205\u0207\u0209\u020B\u020D\u020F\u0211\u0213\u0215\u0217\u0219\u021B\u021D\u021F\u0221\u0223\u0225\u0227\u0229\u022B\u022D\u022F\u0231\u0233-\u0239\u023C\u023F\u0240\u0242\u0247\u0249\u024B\u024D\u024F\u2C61\u2C65\u2C66\u2C68\u2C6A\u2C6C\u2C71\u2C73\u2C74\u2C76-\u2C7B\uA723\uA725\uA727\uA729\uA72B\uA72D\uA72F-\uA731\uA733\uA735\uA737\uA739\uA73B\uA73D\uA73F\uA741\uA743\uA745\uA747\uA749\uA74B\uA74D\uA74F\uA751\uA753\uA755\uA757\uA759\uA75B\uA75D\uA75F\uA761\uA763\uA765\uA767\uA769\uA76B\uA76D\uA76F\uA771-\uA778\uA77A\uA77C\uA77F\uA781\uA783\uA785\uA787\uA78C\uA78E\uA791\uA793-\uA795\uA797\uA799\uA79B\uA79D\uA79F\uA7A1\uA7A3\uA7A5\uA7A7\uA7A9\uA7AF\uA7B5\uA7B7\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E01\u1E03\u1E05\u1E07\u1E09\u1E0B\u1E0D\u1E0F\u1E11\u1E13\u1E15\u1E17\u1E19\u1E1B\u1E1D\u1E1F\u1E21\u1E23\u1E25\u1E27\u1E29\u1E2B\u1E2D\u1E2F\u1E31\u1E33\u1E35\u1E37\u1E39\u1E3B\u1E3D\u1E3F\u1E41\u1E43\u1E45\u1E47\u1E49\u1E4B\u1E4D\u1E4F\u1E51\u1E53\u1E55\u1E57\u1E59\u1E5B\u1E5D\u1E5F\u1E61\u1E63\u1E65\u1E67\u1E69\u1E6B\u1E6D\u1E6F\u1E71\u1E73\u1E75\u1E77\u1E79\u1E7B\u1E7D\u1E7F\u1E81\u1E83\u1E85\u1E87\u1E89\u1E8B\u1E8D\u1E8F\u1E91\u1E93\u1E95-\u1E9D\u1E9F\u1EA1\u1EA3\u1EA5\u1EA7\u1EA9\u1EAB\u1EAD\u1EAF\u1EB1\u1EB3\u1EB5\u1EB7\u1EB9\u1EBB\u1EBD\u1EBF\u1EC1\u1EC3\u1EC5\u1EC7\u1EC9\u1ECB\u1ECD\u1ECF\u1ED1\u1ED3\u1ED5\u1ED7\u1ED9\u1EDB\u1EDD\u1EDF\u1EE1\u1EE3\u1EE5\u1EE7\u1EE9\u1EEB\u1EED\u1EEF\u1EF1\u1EF3\u1EF5\u1EF7\u1EF9\u1EFB\u1EFD\u1EFFёа-яәөүҗңһα-ωάέίόώήύа-щюяіїєґѓѕјљњќѐѝ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F\'"”“`‘´’‚,„»«「」『』()〔〕【】《》〈〉])\.(?=[A-Z\uFF21-\uFF3A\u00C0-\u00D6\u00D8-\u00DE\u0100\u0102\u0104\u0106\u0108\u010A\u010C\u010E\u0110\u0112\u0114\u0116\u0118\u011A\u011C\u011E\u0120\u0122\u0124\u0126\u0128\u012A\u012C\u012E\u0130\u0132\u0134\u0136\u0139\u013B\u013D\u013F\u0141\u0143\u0145\u0147\u014A\u014C\u014E\u0150\u0152\u0154\u0156\u0158\u015A\u015C\u015E\u0160\u0162\u0164\u0166\u0168\u016A\u016C\u016E\u0170\u0172\u0174\u0176\u0178\u0179\u017B\u017D\u0181\u0182\u0184\u0186\u0187\u0189-\u018B\u018E-\u0191\u0193\u0194\u0196-\u0198\u019C\u019D\u019F\u01A0\u01A2\u01A4\u01A6\u01A7\u01A9\u01AC\u01AE\u01AF\u01B1-\u01B3\u01B5\u01B7\u01B8\u01BC\u01C4\u01C7\u01CA\u01CD\u01CF\u01D1\u01D3\u01D5\u01D7\u01D9\u01DB\u01DE\u01E0\u01E2\u01E4\u01E6\u01E8\u01EA\u01EC\u01EE\u01F1\u01F4\u01F6-\u01F8\u01FA\u01FC\u01FE\u0200\u0202\u0204\u0206\u0208\u020A\u020C\u020E\u0210\u0212\u0214\u0216\u0218\u021A\u021C\u021E\u0220\u0222\u0224\u0226\u0228\u022A\u022C\u022E\u0230\u0232\u023A\u023B\u023D\u023E\u0241\u0243-\u0246\u0248\u024A\u024C\u024E\u2C60\u2C62-\u2C64\u2C67\u2C69\u2C6B\u2C6D-\u2C70\u2C72\u2C75\u2C7E\u2C7F\uA722\uA724\uA726\uA728\uA72A\uA72C\uA72E\uA732\uA734\uA736\uA738\uA73A\uA73C\uA73E\uA740\uA742\uA744\uA746\uA748\uA74A\uA74C\uA74E\uA750\uA752\uA754\uA756\uA758\uA75A\uA75C\uA75E\uA760\uA762\uA764\uA766\uA768\uA76A\uA76C\uA76E\uA779\uA77B\uA77D\uA77E\uA780\uA782\uA784\uA786\uA78B\uA78D\uA790\uA792\uA796\uA798\uA79A\uA79C\uA79E\uA7A0\uA7A2\uA7A4\uA7A6\uA7A8\uA7AA-\uA7AE\uA7B0-\uA7B4\uA7B6\uA7B8\u1E00\u1E02\u1E04\u1E06\u1E08\u1E0A\u1E0C\u1E0E\u1E10\u1E12\u1E14\u1E16\u1E18\u1E1A\u1E1C\u1E1E\u1E20\u1E22\u1E24\u1E26\u1E28\u1E2A\u1E2C\u1E2E\u1E30\u1E32\u1E34\u1E36\u1E38\u1E3A\u1E3C\u1E3E\u1E40\u1E42\u1E44\u1E46\u1E48\u1E4A\u1E4C\u1E4E\u1E50\u1E52\u1E54\u1E56\u1E58\u1E5A\u1E5C\u1E5E\u1E60\u1E62\u1E64\u1E66\u1E68\u1E6A\u1E6C\u1E6E\u1E70\u1E72\u1E74\u1E76\u1E78\u1E7A\u1E7C\u1E7E\u1E80\u1E82\u1E84\u1E86\u1E88\u1E8A\u1E8C\u1E8E\u1E90\u1E92\u1E94\u1E9E\u1EA0\u1EA2\u1EA4\u1EA6\u1EA8\u1EAA\u1EAC\u1EAE\u1EB0\u1EB2\u1EB4\u1EB6\u1EB8\u1EBA\u1EBC\u1EBE\u1EC0\u1EC2\u1EC4\u1EC6\u1EC8\u1ECA\u1ECC\u1ECE\u1ED0\u1ED2\u1ED4\u1ED6\u1ED8\u1EDA\u1EDC\u1EDE\u1EE0\u1EE2\u1EE4\u1EE6\u1EE8\u1EEA\u1EEC\u1EEE\u1EF0\u1EF2\u1EF4\u1EF6\u1EF8\u1EFA\u1EFC\u1EFEЁА-ЯӘӨҮҖҢҺΑ-ΩΆΈΊΌΏΉΎА-ЩЮЯІЇЄҐЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F\'"”“`‘´’‚,„»«「」『』()〔〕【】《》〈〉])|(?<=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F]),(?=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F])|(?<=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F0-9])[:<>=/](?=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F])|(?<=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F]['’])(?=[A-Za-z\uFF21-\uFF3A\uFF41-\uFF5A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u017F\u0180-\u01BF\u01C4-\u024F\u2C60-\u2C7B\u2C7E\u2C7F\uA722-\uA76F\uA771-\uA787\uA78B-\uA78E\uA790-\uA7B9\uA7FA\uAB30-\uAB5A\uAB60-\uAB64\u0250-\u02AF\u1D00-\u1D25\u1D6B-\u1D77\u1D79-\u1D9A\u1E00-\u1EFFёа-яЁА-ЯәөүҗңһӘӨҮҖҢҺα-ωάέίόώήύΑ-ΩΆΈΊΌΏΉΎа-щюяіїєґА-ЩЮЯІЇЄҐѓѕјљњќѐѝЃЅЈЉЊЌЀЍ\u1200-\u137F\u0980-\u09FF\u0591-\u05F4\uFB1D-\uFB4F\u0620-\u064A\u066E-\u06D5\u06E5-\u06FF\u0750-\u077F\u08A0-\u08BD\uFB50-\uFBB1\uFBD3-\uFD3D\uFD50-\uFDC7\uFDF0-\uFDFB\uFE70-\uFEFC\U0001EE00-\U0001EEBB\u0D80-\u0DFF\u0900-\u097F\u0C80-\u0CFF\u0B80-\u0BFF\u0C00-\u0C7F\uAC00-\uD7AF\u1100-\u11FF\u4E00-\u62FF\u6300-\u77FF\u7800-\u8CFF\u8D00-\u9FFF\u3400-\u4DBF\U00020000-\U000215FF\U00021600-\U000230FF\U00023100-\U000245FF\U00024600-\U000260FF\U00026100-\U000275FF\U00027600-\U000290FF\U00029100-\U0002A6DF\U0002A700-\U0002B73F\U0002B740-\U0002B81F\U0002B820-\U0002CEAF\U0002CEB0-\U0002EBEF\u2E80-\u2EFF\u2F00-\u2FDF\u2FF0-\u2FFF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\uF900-\uFAFF\uFE30-\uFE4F\U0001F200-\U0001F2FF\U0002F800-\U0002FA1F0-9])|('ls|'l|'ns|'t|'m|'n|-les|-la|-lo|-li|-los|-me|-nos|-te|-vos|-se|-hi|-ne|-ho)(?![A-Za-z])|(-l'|-m'|-t'|-n')�token_match��url_match�
2
+ ��A�
3
+ � ��A� �'��A�'�''��A�''�'l��A�'l�'ls��A�'ls�'m��A�'m�'n��A�'n�'ns��A�'ns�'s��A�'s�'t��A�'t�(*_*)��A�(*_*)�(-8��A�(-8�(-:��A�(-:�(-;��A�(-;�(-_-)��A�(-_-)�(._.)��A�(._.)�(:��A�(:�(;��A�(;�(=��A�(=�(>_<)��A�(>_<)�(^_^)��A�(^_^)�(o:��A�(o:�(¬_¬)��A�(¬_¬)�(ಠ_ಠ)��A�(ಠ_ಠ)�(╯°□°)╯︵┻━┻��A�(╯°□°)╯︵┻━┻�)-:��A�)-:�):��A�):�-_-��A�-_-�-__-��A�-__-�._.��A�._.�0.0��A�0.0�0.o��A�0.o�0_0��A�0_0�0_o��A�0_o�10a.m.��A�10�A�a.m.C�a.m.�10am��A�10�A�amC�a.m.�10p.m.��A�10�A�p.m.C�p.m.�10pm��A�10�A�pmC�p.m.�11a.m.��A�11�A�a.m.C�a.m.�11am��A�11�A�amC�a.m.�11p.m.��A�11�A�p.m.C�p.m.�11pm��A�11�A�pmC�p.m.�12a.m.��A�12�A�a.m.C�a.m.�12am��A�12�A�amC�a.m.�12m.��A�12�A�m.C�p.m.�12p.m.��A�12�A�p.m.C�p.m.�12pm��A�12�A�pmC�p.m.�1a.m.��A�1�A�a.m.C�a.m.�1am��A�1�A�amC�a.m.�1p.m.��A�1�A�p.m.C�p.m.�1pm��A�1�A�pmC�p.m.�2a.m.��A�2�A�a.m.C�a.m.�2am��A�2�A�amC�a.m.�2p.m.��A�2�A�p.m.C�p.m.�2pm��A�2�A�pmC�p.m.�3a.m.��A�3�A�a.m.C�a.m.�3am��A�3�A�amC�a.m.�3p.m.��A�3�A�p.m.C�p.m.�3pm��A�3�A�pmC�p.m.�4a.m.��A�4�A�a.m.C�a.m.�4am��A�4�A�amC�a.m.�4p.m.��A�4�A�p.m.C�p.m.�4pm��A�4�A�pmC�p.m.�5a.m.��A�5�A�a.m.C�a.m.�5am��A�5�A�amC�a.m.�5p.m.��A�5�A�p.m.C�p.m.�5pm��A�5�A�pmC�p.m.�6a.m.��A�6�A�a.m.C�a.m.�6am��A�6�A�amC�a.m.�6p.m.��A�6�A�p.m.C�p.m.�6pm��A�6�A�pmC�p.m.�7a.m.��A�7�A�a.m.C�a.m.�7am��A�7�A�amC�a.m.�7p.m.��A�7�A�p.m.C�p.m.�7pm��A�7�A�pmC�p.m.�8)��A�8)�8-)��A�8-)�8-D��A�8-D�8D��A�8D�8a.m.��A�8�A�a.m.C�a.m.�8am��A�8�A�amC�a.m.�8p.m.��A�8�A�p.m.C�p.m.�8pm��A�8�A�pmC�p.m.�9a.m.��A�9�A�a.m.C�a.m.�9am��A�9�A�amC�a.m.�9p.m.��A�9�A�p.m.C�p.m.�9pm��A�9�A�pmC�p.m.�:'(��A�:'(�:')��A�:')�:'-(��A�:'-(�:'-)��A�:'-)�:(��A�:(�:((��A�:((�:(((��A�:(((�:()��A�:()�:)��A�:)�:))��A�:))�:)))��A�:)))�:*��A�:*�:-(��A�:-(�:-((��A�:-((�:-(((��A�:-(((�:-)��A�:-)�:-))��A�:-))�:-)))��A�:-)))�:-*��A�:-*�:-/��A�:-/�:-0��A�:-0�:-3��A�:-3�:->��A�:->�:-D��A�:-D�:-O��A�:-O�:-P��A�:-P�:-X��A�:-X�:-]��A�:-]�:-o��A�:-o�:-p��A�:-p�:-x��A�:-x�:-|��A�:-|�:-}��A�:-}�:/��A�:/�:0��A�:0�:1��A�:1�:3��A�:3�:>��A�:>�:D��A�:D�:O��A�:O�:P��A�:P�:X��A�:X�:]��A�:]�:o��A�:o�:o)��A�:o)�:p��A�:p�:x��A�:x�:|��A�:|�:}��A�:}�:’(��A�:’(�:’)��A�:’)�:’-(��A�:’-(�:’-)��A�:’-)�;)��A�;)�;-)��A�;-)�;-D��A�;-D�;D��A�;D�;_;��A�;_;�<.<��A�<.<�</3��A�</3�<3��A�<3�<33��A�<33�<333��A�<333�<space>��A�<space>�=(��A�=(�=)��A�=)�=/��A�=/�=3��A�=3�=D��A�=D�=[��A�=[�=]��A�=]�=|��A�=|�>.<��A�>.<�>.>��A�>.>�>:(��A�>:(�>:o��A�>:o�><(((*>��A�><(((*>�@_@��A�@_@�C++��A�C++�Dr.��A�Dr.C�doctor�Dra.��A�Dra.C�doctora�O.O��A�O.O�O.o��A�O.o�O_O��A�O_O�O_o��A�O_o�Sr.��A�Sr.C�senyor�Sra.��A�Sra.C�senyora�Srta.��A�Srta.C�senyoreta�St.��A�St.C�sant�Sta.��A�Sta.C�santa�V.V��A�V.V�V_V��A�V_V�XD��A�XD�XDD��A�XDD�[-:��A�[-:�[:��A�[:�[=��A�[=�\")��A�\")�\n��A�\n�\t��A�\t�]=��A�]=�^_^��A�^_^�^__^��A�^__^�^___^��A�^___^�a.��A�a.�abr.��A�abr.C�abril�al��A�a�A�lC�el�als��A�a�A�lsC�els�aprox.��A�aprox.C�aproximadament�b.��A�b.�c.��A�c.�d.��A�d.�dec.��A�dec.C�desembre�del��A�dC�de�A�el�dels��A�dC�de�A�els�e.��A�e.�f.��A�f.�feb.��A�feb.C�febrer�g.��A�g.�gen.��A�gen.C�gener�h.��A�h.�holahola��A�holaholaC�cocacola�i.��A�i.�j.��A�j.�jul.��A�jul.C�juliol�k.��A�k.�l.��A�l.�m.��A�m.�n.��A�n.�nov.��A�nov.C�novembre�núm��A�númC�número�o.��A�o.�o.0��A�o.0�o.O��A�o.O�o.o��A�o.o�o_0��A�o_0�o_O��A�o_O�o_o��A�o_o�oct.��A�oct.C�octubre�p.��A�p.�p.ex.��A�p.ex.C�per exemple�pel��A�pC�per�A�el�pels��A�pC�per�A�els�pl.��A�pl.C�plaça�pàg.��A�pàg.C�pàgina�q.��A�q.�r.��A�r.�s.��A�s.�set.��A�set.C�setembre�t.��A�t.�u.��A�u.�v.��A�v.�v.v��A�v.v�v_v��A�v_v�w.��A�w.�x.��A�x.�xD��A�xD�xDD��A�xDD�y.��A�y.�z.��A�z.� ��A� C� �¯\(ツ)/¯��A�¯\(ツ)/¯�°C.��A�°�A�C�A�.�°F.��A�°�A�F�A�.�°K.��A�°�A�K�A�.�°c.��A�°�A�c�A�.�°f.��A�°�A�f�A�.�°k.��A�°�A�k�A�.�à.��A�à.�ä.��A�ä.�è.��A�è.�é.��A�é.�í.��A�í.�ò.��A�ò.�ó.��A�ó.�ö.��A�ö.�ú.��A�ú.�ü.��A�ü.�ಠ_ಠ��A�ಠ_ಠ�ಠ︵ಠ��A�ಠ︵ಠ�—��A�—�’��A�’�’l��A�’l�’ls��A�’ls�’m��A�’m�’n��A�’n�’ns��A�’ns�’s��A�’s�’t��A�’t�’’��A�’’
vocab/key2row ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:843c932e2e359197afe1a8a87e37dee491be10f6520a729d2ecb3ce3dfa70e86
3
+ size 6868193
vocab/lookups.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76be8b528d0075f7aae98d6fa57a6d3c83ae480a8469e668d7b0af968995ac71
3
+ size 1
vocab/strings.json ADDED
The diff for this file is too large to render. See raw diff
 
vocab/vectors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:507f52fbbcad183c891c3ba15a60566e7a5e5a57aa618b2ee705851081828210
3
+ size 600000128
vocab/vectors.cfg ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "mode":"default"
3
+ }