Hugging Face's logo

Datasets:
clarin-knext
/
wsd_plwordnet_glex

Tasks:
Token Classification
Sub-tasks:
word-sense-disambiguation
Languages:
Polish
Size:
1M<n<10M
License:
Dataset card Files
xet
 Community
4
Arkany commited on
Commit
9f17960
·
1 Parent(s): d5228f5

Update wsd_plwordnet_glex.py

Browse files

Fix loading for usage_examples

Files changed (1) hide show
  1. wsd_plwordnet_glex.py +68 -9
wsd_plwordnet_glex.py CHANGED
@@ -42,30 +42,78 @@ class WsdPlwordnetGlex(datasets.GeneratorBasedBuilder):
42
 
43
  def _info(self):
44
  features = datasets.Features()
45
- if self.config.name == "definitions" or self.config.name == "usage_examples":
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
46
  features = datasets.Features(
47
  {
48
  "pl_sense": datasets.Value("string"),
49
  "text": datasets.Value("string"),
50
- "tokens": datasets.Value("string"),
51
- "phrases": datasets.Value("string"),
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
  "plWN_lex_id": datasets.Value("string"),
53
  "plWN_syn_id": datasets.Value("string"),
54
- "wsd": datasets.Value("string"),
55
  }
56
  )
57
  if self.config.name == "lemma_candidates":
58
  features = datasets.Features(
59
  {
60
  "lemma": datasets.Value("string"),
61
- "candidates": datasets.Value("string"),
62
  }
63
  )
64
  return datasets.DatasetInfo(
65
  description=_DESCRIPTION,
66
  features=features,
67
  )
68
-
69
  def _split_generators(self, dl_manager):
70
  dl_dir = dl_manager.download_and_extract(self.config.data_dir)
71
  return [
@@ -79,7 +127,18 @@ class WsdPlwordnetGlex(datasets.GeneratorBasedBuilder):
79
  """Generate examples."""
80
 
81
  with open(path, encoding="utf-8") as f:
82
- if self.config.name == "definitions" or self.config.name == "usage_examples":
 
 
 
 
 
 
 
 
 
 
 
83
  for k,line in enumerate(f):
84
  data = json.loads(line)
85
  yield k, {
@@ -89,10 +148,10 @@ class WsdPlwordnetGlex(datasets.GeneratorBasedBuilder):
89
  "phrases": data["phrases"],
90
  "plWN_lex_id": data["plWN_lex_id"],
91
  "plWN_syn_id": data["plWN_syn_id"],
92
- "wsd": data["wsd"],
93
  }
94
 
95
  if self.config.name == "lemma_candidates":
96
  data = json.loads(f.read())
97
  processed = ((ind,{"lemma": k, "candidates": v}) for ind,(k,v) in enumerate(data.items()))
98
- yield from processed
 
42
 
43
  def _info(self):
44
  features = datasets.Features()
45
+ if self.config.name == "usage_examples":
46
+ features = features = datasets.Features(
47
+ {
48
+
49
+ "text": datasets.Value("string"),
50
+ "tokens": datasets.features.Sequence(
51
+ dict(
52
+ {
53
+ "index": datasets.Value("int32"),
54
+ "position": datasets.features.Sequence(
55
+ length=2,
56
+ feature=datasets.Value("int32"),
57
+ ),
58
+ "orth": datasets.Value("string"),
59
+ "lemma": datasets.Value("string"),
60
+ "pos": datasets.Value("string"),
61
+ "ctag": datasets.Value("string"),
62
+ }
63
+ ),
64
+ ),
65
+ "phrases": datasets.features.Sequence(datasets.Value("string")),
66
+ "wsd": datasets.features.Sequence(
67
+ dict({
68
+ "pl_sense": datasets.Value("string"),
69
+ "plWN_lex_id": datasets.Value("string"),
70
+ "plWN_syn_id": datasets.Value("string"),
71
+ "plWN_lex_legacy_id": datasets.Value("string"),
72
+ "plWN_syn_legacy_id": datasets.Value("string"),
73
+ "index": datasets.Value("int32"),
74
+ })
75
+ ),
76
+ 'context_file': datasets.Value('string'),
77
+ }
78
+ )
79
+ if self.config.name == "definitions":
80
  features = datasets.Features(
81
  {
82
  "pl_sense": datasets.Value("string"),
83
  "text": datasets.Value("string"),
84
+ "tokens": datasets.features.Sequence(
85
+ dict(
86
+ {
87
+ "index": datasets.Value("int32"),
88
+ "position": datasets.features.Sequence(
89
+ length=2,
90
+ feature=datasets.Value("int32"),
91
+ ),
92
+ "orth": datasets.Value("string"),
93
+ "lemma": datasets.Value("string"),
94
+ "pos": datasets.Value("string"),
95
+ "ctag": datasets.Value("string"),
96
+ }
97
+ ),
98
+ ),
99
+ "phrases": datasets.features.Sequence(datasets.Value("string")),
100
  "plWN_lex_id": datasets.Value("string"),
101
  "plWN_syn_id": datasets.Value("string"),
102
+ "wsd_indices": datasets.features.Sequence(datasets.Value("int32")),
103
  }
104
  )
105
  if self.config.name == "lemma_candidates":
106
  features = datasets.Features(
107
  {
108
  "lemma": datasets.Value("string"),
109
+ "candidates": datasets.features.Sequence(datasets.Value("string")),
110
  }
111
  )
112
  return datasets.DatasetInfo(
113
  description=_DESCRIPTION,
114
  features=features,
115
  )
116
+
117
  def _split_generators(self, dl_manager):
118
  dl_dir = dl_manager.download_and_extract(self.config.data_dir)
119
  return [
 
127
  """Generate examples."""
128
 
129
  with open(path, encoding="utf-8") as f:
130
+ if self.config.name == "usage_examples":
131
+ for k,line in enumerate(f):
132
+ data = json.loads(line)
133
+ yield k, {
134
+ "text": data["text"],
135
+ "tokens": data["tokens"],
136
+ "phrases": data["phrases"],
137
+ "wsd": data["wsd"],
138
+ 'context_file': data['context_file'],
139
+ }
140
+
141
+ if self.config.name == "definitions":
142
  for k,line in enumerate(f):
143
  data = json.loads(line)
144
  yield k, {
 
148
  "phrases": data["phrases"],
149
  "plWN_lex_id": data["plWN_lex_id"],
150
  "plWN_syn_id": data["plWN_syn_id"],
151
+ "wsd_indices": data["wsd_indices"],
152
  }
153
 
154
  if self.config.name == "lemma_candidates":
155
  data = json.loads(f.read())
156
  processed = ((ind,{"lemma": k, "candidates": v}) for ind,(k,v) in enumerate(data.items()))
157
+ yield from processed