aembryonic commited on
Commit
e3342b7
1 Parent(s): c8de64c

add version 2

Browse files
Files changed (1) hide show
  1. humset.py +49 -34
humset.py CHANGED
@@ -47,25 +47,26 @@ _URLs = {
47
  "train": "data/train.jsonl",
48
  "dev": "data/validation.jsonl",
49
  "test": "data/test.jsonl",
 
 
 
 
 
50
  }
51
  }
52
 
53
 
54
  _SUPPORTED_VERSIONS = [
55
- # Using cased version.
56
  datasets.Version("1.0.0", "Only primary tags"),
57
- # Same data as 0.0.2
58
- #datasets.Version("1.0.0", ""),
59
- # Having the model predict newline separators makes it easier to evaluate
60
- # using summary-level ROUGE.
61
- #datasets.Version("2.0.0", "Separate target sentences with newline."),
62
  ]
63
 
64
-
65
- DEFAULT_CONFIG_NAME = "1.0.0" #datasets.Version("1.0", "Using cased version.")
66
 
67
  ## COMMON FEATURES
68
- MAIN_FEATURES = datasets.Features(
69
  {
70
  "entry_id": datasets.Value("string"),
71
  "lead_id": datasets.Value("string"),
@@ -186,6 +187,34 @@ MAIN_FEATURES = datasets.Features(
186
  )
187
 
188
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
189
  class HumsetConfig(datasets.BuilderConfig):
190
  """BuilderConfig for DuoRC SelfRC."""
191
 
@@ -198,7 +227,6 @@ class HumsetConfig(datasets.BuilderConfig):
198
 
199
 
200
  class Humset(datasets.GeneratorBasedBuilder):
201
- """DuoRC Dataset"""
202
 
203
  BUILDER_CONFIGS = [
204
  HumsetConfig(
@@ -216,7 +244,17 @@ class Humset(datasets.GeneratorBasedBuilder):
216
  # This is the description that will appear on the datasets page.
217
  description=_DESCRIPTION,
218
  # This defines the different columns of the dataset and their types
219
- features=MAIN_FEATURES,
 
 
 
 
 
 
 
 
 
 
220
  homepage=_HOMEPAGE,
221
  license=_LICENSE,
222
  citation=_CITATION,
@@ -260,26 +298,3 @@ class Humset(datasets.GeneratorBasedBuilder):
260
  if self.config.name == "1.0.0":
261
  yield idx, row
262
  idx+=1
263
- #for example in duorc:
264
-
265
-
266
- """
267
- plot_id = example["id"]
268
- plot = example["plot"].strip()
269
- title = example["title"].strip()
270
- for qas in example["qa"]:
271
- question_id = qas["id"]
272
- question = qas["question"].strip()
273
- answers = [answer.strip() for answer in qas["answers"]]
274
- no_answer = qas["no_answer"]
275
-
276
- yield question_id, {
277
- "title": title,
278
- "plot": plot,
279
- "question": question,
280
- "plot_id": plot_id,
281
- "question_id": question_id,
282
- "answers": answers,
283
- "no_answer": no_answer,
284
- }
285
- """
 
47
  "train": "data/train.jsonl",
48
  "dev": "data/validation.jsonl",
49
  "test": "data/test.jsonl",
50
+ },
51
+ "2.0.0": {
52
+ "train": "data/train_1_1.jsonl",
53
+ "dev": "data/validation_1_1_.jsonl",
54
+ "test": "data/test_1_1.jsonl"
55
  }
56
  }
57
 
58
 
59
  _SUPPORTED_VERSIONS = [
60
+ # First version
61
  datasets.Version("1.0.0", "Only primary tags"),
62
+ # Second version
63
+ datasets.Version("2.0.0", "Extented data points including secondary tags and geolocations"),
 
 
 
64
  ]
65
 
66
+ DEFAULT_CONFIG_NAME = "2.0.0"
 
67
 
68
  ## COMMON FEATURES
69
+ FIRST_FEATURES = datasets.Features(
70
  {
71
  "entry_id": datasets.Value("string"),
72
  "lead_id": datasets.Value("string"),
 
187
  )
188
 
189
 
190
+ SECOND_FEATURES = datasets.Features(
191
+ {
192
+ "entry_id": datasets.Value("string"),
193
+ "lead_id": datasets.Value("string"),
194
+ "lang": datasets.Value("string"),
195
+ "n_tokens": datasets.Value("int64"),
196
+ "project_title": datasets.Value("string"),
197
+ "created_at": datasets.Value("string"),
198
+ "document": datasets.Value("string"),
199
+ "source_title": datasets.Value("string"),
200
+ "author_title": datasets.Value("string"),
201
+ "excerpt": datasets.Value("string"),
202
+ "geo_location": datasets.Sequence(datasets.Value("string"), length=-1),
203
+ "sectors": datasets.Sequence(datasets.Value("string"), length=-1),
204
+ "pillars_1d": datasets.Sequence(datasets.Value("string"), length=-1),
205
+ "pillars_2d": datasets.Sequence(datasets.Value("string"), length=-1),
206
+ "subpillars_1d": datasets.Sequence(datasets.Value("string"), length=-1),
207
+ "subpillars_2d": datasets.Sequence(datasets.Value("string"), length=-1),
208
+ "displaced": datasets.Sequence(datasets.Value("string"), length=-1),
209
+ "non_displaced": datasets.Sequence(datasets.Value("string"), length=-1),
210
+ "affected": datasets.Sequence(datasets.Value("string"), length=-1),
211
+ "severity": datasets.Sequence(datasets.Value("string"), length=-1),
212
+ "age": datasets.Sequence(datasets.Value("string"), length=-1),
213
+ "gender": datasets.Sequence(datasets.Value("string"), length=-1),
214
+ "specific_needs_groups": datasets.Sequence(datasets.Value("string"), length=-1)
215
+ }
216
+ )
217
+
218
  class HumsetConfig(datasets.BuilderConfig):
219
  """BuilderConfig for DuoRC SelfRC."""
220
 
 
227
 
228
 
229
  class Humset(datasets.GeneratorBasedBuilder):
 
230
 
231
  BUILDER_CONFIGS = [
232
  HumsetConfig(
 
244
  # This is the description that will appear on the datasets page.
245
  description=_DESCRIPTION,
246
  # This defines the different columns of the dataset and their types
247
+ features=FIRST_FEATURES,
248
+ homepage=_HOMEPAGE,
249
+ license=_LICENSE,
250
+ citation=_CITATION,
251
+ )
252
+ elif self.config.name == "2.0.0":
253
+ return datasets.DatasetInfo(
254
+ # This is the description that will appear on the datasets page.
255
+ description=_DESCRIPTION,
256
+ # This defines the different columns of the dataset and their types
257
+ features=SECOND_FEATURES,
258
  homepage=_HOMEPAGE,
259
  license=_LICENSE,
260
  citation=_CITATION,
 
298
  if self.config.name == "1.0.0":
299
  yield idx, row
300
  idx+=1