holylovenia commited on
Commit
f811fcb
·
verified ·
1 Parent(s): 5cd446b

Upload id_hsd_nofaaulia.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. id_hsd_nofaaulia.py +13 -13
id_hsd_nofaaulia.py CHANGED
@@ -21,9 +21,9 @@ import datasets
21
  import pandas as pd
22
  from sklearn.model_selection import train_test_split
23
 
24
- from nusacrowd.utils import schemas
25
- from nusacrowd.utils.configs import NusantaraConfig
26
- from nusacrowd.utils.constants import (DEFAULT_NUSANTARA_VIEW_NAME,
27
  DEFAULT_SOURCE_VIEW_NAME, Tasks)
28
 
29
  _CITATION = """\
@@ -64,27 +64,27 @@ _URLS = {
64
  _SUPPORTED_TASKS = [Tasks.SENTIMENT_ANALYSIS]
65
 
66
  _SOURCE_VERSION = "1.0.0"
67
- _NUSANTARA_VERSION = "1.0.0"
68
 
69
  class IdHSDNofaaulia(datasets.GeneratorBasedBuilder):
70
  """Indonesian hate speech detection for long article."""
71
 
72
  SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
73
- NUSANTARA_VERSION = datasets.Version(_NUSANTARA_VERSION)
74
 
75
  BUILDER_CONFIGS = [
76
- NusantaraConfig(
77
  name="id_hsd_nofaaulia_source",
78
  version=SOURCE_VERSION,
79
  description="id_hsd_nofaaulia source schema",
80
  schema="source",
81
  subset_id="id_hsd_nofaaulia",
82
  ),
83
- NusantaraConfig(
84
- name="id_hsd_nofaaulia_nusantara_text",
85
- version=NUSANTARA_VERSION,
86
  description="id_hsd_nofaaulia Nusantara schema",
87
- schema="nusantara_text",
88
  subset_id="id_hsd_nofaaulia",
89
  ),
90
  ]
@@ -95,7 +95,7 @@ class IdHSDNofaaulia(datasets.GeneratorBasedBuilder):
95
 
96
  if self.config.schema == "source":
97
  features = datasets.Features({"index": datasets.Value("string"), "text": datasets.Value("string"), "label": datasets.Value("string")})
98
- elif self.config.schema == "nusantara_text":
99
  features = schemas.text_features(["0", "1"])
100
 
101
  return datasets.DatasetInfo(
@@ -121,7 +121,7 @@ class IdHSDNofaaulia(datasets.GeneratorBasedBuilder):
121
  target = "label"
122
 
123
  # The split follows the implementation below
124
- # https://github.com/IndoNLP/nusa-crowd/blob/master/nusantara/utils/schemas/pairs.py
125
  # test_size=0.1, random_state=42
126
  # tested locally using :
127
  # scikit-learn 1.1.2
@@ -163,7 +163,7 @@ class IdHSDNofaaulia(datasets.GeneratorBasedBuilder):
163
  }
164
  yield row.Index, ex
165
 
166
- elif self.config.schema == "nusantara_text":
167
  for row in df.itertuples():
168
  ex = {
169
  "id": str(row.Index),
 
21
  import pandas as pd
22
  from sklearn.model_selection import train_test_split
23
 
24
+ from seacrowd.utils import schemas
25
+ from seacrowd.utils.configs import SEACrowdConfig
26
+ from seacrowd.utils.constants import (DEFAULT_SEACROWD_VIEW_NAME,
27
  DEFAULT_SOURCE_VIEW_NAME, Tasks)
28
 
29
  _CITATION = """\
 
64
  _SUPPORTED_TASKS = [Tasks.SENTIMENT_ANALYSIS]
65
 
66
  _SOURCE_VERSION = "1.0.0"
67
+ _SEACROWD_VERSION = "2024.06.20"
68
 
69
  class IdHSDNofaaulia(datasets.GeneratorBasedBuilder):
70
  """Indonesian hate speech detection for long article."""
71
 
72
  SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
73
+ SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
74
 
75
  BUILDER_CONFIGS = [
76
+ SEACrowdConfig(
77
  name="id_hsd_nofaaulia_source",
78
  version=SOURCE_VERSION,
79
  description="id_hsd_nofaaulia source schema",
80
  schema="source",
81
  subset_id="id_hsd_nofaaulia",
82
  ),
83
+ SEACrowdConfig(
84
+ name="id_hsd_nofaaulia_seacrowd_text",
85
+ version=SEACROWD_VERSION,
86
  description="id_hsd_nofaaulia Nusantara schema",
87
+ schema="seacrowd_text",
88
  subset_id="id_hsd_nofaaulia",
89
  ),
90
  ]
 
95
 
96
  if self.config.schema == "source":
97
  features = datasets.Features({"index": datasets.Value("string"), "text": datasets.Value("string"), "label": datasets.Value("string")})
98
+ elif self.config.schema == "seacrowd_text":
99
  features = schemas.text_features(["0", "1"])
100
 
101
  return datasets.DatasetInfo(
 
121
  target = "label"
122
 
123
  # The split follows the implementation below
124
+ # https://github.com/IndoNLP/nusa-crowd/blob/master/seacrowd/utils/schemas/pairs.py
125
  # test_size=0.1, random_state=42
126
  # tested locally using :
127
  # scikit-learn 1.1.2
 
163
  }
164
  yield row.Index, ex
165
 
166
+ elif self.config.schema == "seacrowd_text":
167
  for row in df.itertuples():
168
  ex = {
169
  "id": str(row.Index),