ramybaly commited on
Commit
9ada35a
·
1 Parent(s): ad5cea2

fixed bug in dataset script

Browse files
Files changed (6) hide show
  1. id2coarse_tags.json +2 -2
  2. nerd.py +74 -74
  3. prepare_data.py +9 -6
  4. test.txt +2 -2
  5. train.txt +2 -2
  6. validation.txt +2 -2
id2coarse_tags.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f8ea0be9d72f6d12a799687307099f1b742055095d8f7424ec46c4781769dfc
3
- size 136
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a8f2b723f940c3bcfc2422fedd824fbd92f0126f4f7f38866edb0b376bbaa29
3
+ size 134
nerd.py CHANGED
@@ -70,82 +70,82 @@ class NERD(datasets.GeneratorBasedBuilder):
70
  "tokens": datasets.Sequence(datasets.Value("string")),
71
  "ner_tags": datasets.Sequence(
72
  datasets.features.ClassLabel(
73
- names=['O', 'I-art', 'I-building', 'I-event', 'I-location', 'I-organization', 'I-other', 'I-person', 'I-product']
74
- )
75
- ),
76
- "ner_tags_fine": datasets.Sequence(
77
- datasets.features.ClassLabel(
78
- names=[
79
- 'O',
80
- 'I-art-broadcastprogram',
81
- 'I-art-film',
82
- 'I-art-music',
83
- 'I-art-other',
84
- 'I-art-painting',
85
- 'I-art-writtenart',
86
- 'I-building-airport',
87
- 'I-building-hospital',
88
- 'I-building-hotel',
89
- 'I-building-library',
90
- 'I-building-other',
91
- 'I-building-restaurant',
92
- 'I-building-sportsfacility',
93
- 'I-building-theater',
94
- 'I-event-attack/battle/war/militaryconflict',
95
- 'I-event-disaster',
96
- 'I-event-election',
97
- 'I-event-other',
98
- 'I-event-protest',
99
- 'I-event-sportsevent',
100
- 'I-location-GPE',
101
- 'I-location-bodiesofwater',
102
- 'I-location-island',
103
- 'I-location-mountain',
104
- 'I-location-other',
105
- 'I-location-park',
106
- 'I-location-road/railway/highway/transit',
107
- 'I-organization-company',
108
- 'I-organization-education',
109
- 'I-organization-government/governmentagency',
110
- 'I-organization-media/newspaper',
111
- 'I-organization-other',
112
- 'I-organization-politicalparty',
113
- 'I-organization-religion',
114
- 'I-organization-showorganization',
115
- 'I-organization-sportsleague',
116
- 'I-organization-sportsteam',
117
- 'I-other-astronomything',
118
- 'I-other-award',
119
- 'I-other-biologything',
120
- 'I-other-chemicalthing',
121
- 'I-other-currency',
122
- 'I-other-disease',
123
- 'I-other-educationaldegree',
124
- 'I-other-god',
125
- 'I-other-language',
126
- 'I-other-law',
127
- 'I-other-livingthing',
128
- 'I-other-medical',
129
- 'I-person-actor',
130
- 'I-person-artist/author',
131
- 'I-person-athlete',
132
- 'I-person-director',
133
- 'I-person-other',
134
- 'I-person-politician',
135
- 'I-person-scholar',
136
- 'I-person-soldier',
137
- 'I-product-airplane',
138
- 'I-product-car',
139
- 'I-product-food',
140
- 'I-product-game',
141
- 'I-product-other',
142
- 'I-product-ship',
143
- 'I-product-software',
144
- 'I-product-train',
145
- 'I-product-weapon'
146
- ]
147
  )
148
  ),
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
149
  }
150
  ),
151
  supervised_keys=None,
 
70
  "tokens": datasets.Sequence(datasets.Value("string")),
71
  "ner_tags": datasets.Sequence(
72
  datasets.features.ClassLabel(
73
+ names=['O', 'I-ART', 'I-BUILDING', 'I-EVENT', 'I-LOC', 'I-ORG', 'I-MISC', 'I-PER', 'I-PRODUCT']
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
74
  )
75
  ),
76
+ # "ner_tags_fine": datasets.Sequence(
77
+ # datasets.features.ClassLabel(
78
+ # names=[
79
+ # 'O',
80
+ # 'I-art-broadcastprogram',
81
+ # 'I-art-film',
82
+ # 'I-art-music',
83
+ # 'I-art-other',
84
+ # 'I-art-painting',
85
+ # 'I-art-writtenart',
86
+ # 'I-building-airport',
87
+ # 'I-building-hospital',
88
+ # 'I-building-hotel',
89
+ # 'I-building-library',
90
+ # 'I-building-other',
91
+ # 'I-building-restaurant',
92
+ # 'I-building-sportsfacility',
93
+ # 'I-building-theater',
94
+ # 'I-event-attack/battle/war/militaryconflict',
95
+ # 'I-event-disaster',
96
+ # 'I-event-election',
97
+ # 'I-event-other',
98
+ # 'I-event-protest',
99
+ # 'I-event-sportsevent',
100
+ # 'I-location-GPE',
101
+ # 'I-location-bodiesofwater',
102
+ # 'I-location-island',
103
+ # 'I-location-mountain',
104
+ # 'I-location-other',
105
+ # 'I-location-park',
106
+ # 'I-location-road/railway/highway/transit',
107
+ # 'I-organization-company',
108
+ # 'I-organization-education',
109
+ # 'I-organization-government/governmentagency',
110
+ # 'I-organization-media/newspaper',
111
+ # 'I-organization-other',
112
+ # 'I-organization-politicalparty',
113
+ # 'I-organization-religion',
114
+ # 'I-organization-showorganization',
115
+ # 'I-organization-sportsleague',
116
+ # 'I-organization-sportsteam',
117
+ # 'I-other-astronomything',
118
+ # 'I-other-award',
119
+ # 'I-other-biologything',
120
+ # 'I-other-chemicalthing',
121
+ # 'I-other-currency',
122
+ # 'I-other-disease',
123
+ # 'I-other-educationaldegree',
124
+ # 'I-other-god',
125
+ # 'I-other-language',
126
+ # 'I-other-law',
127
+ # 'I-other-livingthing',
128
+ # 'I-other-medical',
129
+ # 'I-person-actor',
130
+ # 'I-person-artist/author',
131
+ # 'I-person-athlete',
132
+ # 'I-person-director',
133
+ # 'I-person-other',
134
+ # 'I-person-politician',
135
+ # 'I-person-scholar',
136
+ # 'I-person-soldier',
137
+ # 'I-product-airplane',
138
+ # 'I-product-car',
139
+ # 'I-product-food',
140
+ # 'I-product-game',
141
+ # 'I-product-other',
142
+ # 'I-product-ship',
143
+ # 'I-product-software',
144
+ # 'I-product-train',
145
+ # 'I-product-weapon'
146
+ # ]
147
+ # )
148
+ # ),
149
  }
150
  ),
151
  supervised_keys=None,
prepare_data.py CHANGED
@@ -5,7 +5,7 @@ import pickle
5
 
6
  def main():
7
  id2label = {int(i): tag for i, tag in json.load(open('./id2coarse_tags.json', 'r')).items()}
8
- id2label_fine = {int(i): tag for i, tag in json.load(open('./id2fine_tags.json', 'r')).items()}
9
 
10
  files = glob.glob('./*json')
11
 
@@ -23,12 +23,15 @@ def main():
23
  ner_tags_ids = sentence['coarse_tags']
24
  ner_tags_fine_ids = sentence['fine_tags']
25
 
26
- ner_tags = ['-'.join(['I', id2label[t]]) if id2label[t] != 'O' else id2label[t] for t in ner_tags_ids]
27
- ner_tags_fine = ['-'.join(['I', id2label_fine[t]]) if id2label_fine[t] != 'O' else id2label[t] for t in ner_tags_fine_ids]
28
-
29
- for token, tag, tag_fine in zip(tokens, ner_tags, ner_tags_fine):
30
- f.write(f'{token}\t{tag}\t{tag_fine}\n')
31
 
 
 
 
 
 
 
32
  f.write('\n')
33
 
34
 
 
5
 
6
  def main():
7
  id2label = {int(i): tag for i, tag in json.load(open('./id2coarse_tags.json', 'r')).items()}
8
+ # id2label_fine = {int(i): tag for i, tag in json.load(open('./id2fine_tags.json', 'r')).items()}
9
 
10
  files = glob.glob('./*json')
11
 
 
23
  ner_tags_ids = sentence['coarse_tags']
24
  ner_tags_fine_ids = sentence['fine_tags']
25
 
26
+ ner_tags = [id2label[t] for t in ner_tags_ids]
27
+ # ner_tags_fine = ['-'.join(['I', id2label_fine[t]]) if id2label_fine[t] != 'O' else id2label[t] for t in ner_tags_fine_ids]
 
 
 
28
 
29
+ # for token, tag, tag_fine in zip(tokens, ner_tags, ner_tags_fine):
30
+ # f.write(f'{token}\t{tag}\t{tag_fine}\n')
31
+
32
+ for token, tag in zip(tokens, ner_tags):
33
+ f.write(f'{token}\t{tag}\n')
34
+
35
  f.write('\n')
36
 
37
 
test.txt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b1b5ec164a8800c9a70de28f38137780989db7f0f331a7d049bb85c0d3235f85
3
- size 13975793
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc3d935eb96b72f8f0014de6e21467fbdbaad1675ae6e56189acd0b2db0185a
3
+ size 7878894
train.txt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:28a16e4eaa1c2caf717db49467ac91c222825ac231d8e662c053b913600cbe35
3
- size 49030821
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4dba55486479f62f0f7e7150f046edc532effb4e91a9570586cbc31af04e074b
3
+ size 27615430
validation.txt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b319ef76a7051b3447680d826c020de1a78f30dfc1285c1c769d815a0b731033
3
- size 7026135
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9201e711390f19a0301eb0a36b690aa13b68e3e476052f15cd5ef9e4cfa79a45
3
+ size 3960928