ToluClassics commited on
Commit
bbcd3c1
·
1 Parent(s): 20a75c8

modify file to enable different splits per language

Browse files
Files changed (1) hide show
  1. afriqa-gold-passages.py +26 -6
afriqa-gold-passages.py CHANGED
@@ -29,11 +29,23 @@ _LANG_2_PIVOT = {
29
  "kin": "en",
30
  "swa": "en",
31
  "twi": "en",
32
- "wol": "fr",
33
  "yor": "en",
34
  "zul": "en",
35
  }
36
 
 
 
 
 
 
 
 
 
 
 
 
 
 
37
  class AfriQAConfig(datasets.BuilderConfig):
38
  """BuilderConfig for AfriQA"""
39
 
@@ -90,11 +102,19 @@ class AfriQA(datasets.GeneratorBasedBuilder):
90
  }
91
  downloaded_files = dl_manager.download_and_extract(urls_to_download)
92
 
93
- return [
94
- datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}),
95
- datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["dev"]}),
96
- datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test"]}),
97
- ]
 
 
 
 
 
 
 
 
98
 
99
  def _generate_examples(self, filepath):
100
  """Yields examples."""
 
29
  "kin": "en",
30
  "swa": "en",
31
  "twi": "en",
 
32
  "yor": "en",
33
  "zul": "en",
34
  }
35
 
36
+ _LANG_2_SPLITS = {
37
+ "bem": ["train", "dev", "test"],
38
+ "fon": ["train", "dev", "test"],
39
+ "hau": ["train", "dev", "test"],
40
+ "ibo": ["train", "dev", "test"],
41
+ "kin": ["train", "dev", "test"],
42
+ "swa": ["test"],
43
+ "twi": ["train", "dev", "test"],
44
+ "yor": ["train", "test"],
45
+ "zul": ["train", "dev", "test"],
46
+
47
+ }
48
+
49
  class AfriQAConfig(datasets.BuilderConfig):
50
  """BuilderConfig for AfriQA"""
51
 
 
102
  }
103
  downloaded_files = dl_manager.download_and_extract(urls_to_download)
104
 
105
+ splits_list = []
106
+
107
+ for split in _LANG_2_SPLITS[self.config.name]:
108
+
109
+ if split == "train":
110
+ splits_list.append(datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}))
111
+ elif split == "dev":
112
+ splits_list.append(datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["dev"]}))
113
+ elif split == "test":
114
+ splits_list.append(datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test"]}))
115
+
116
+
117
+ return splits_list
118
 
119
  def _generate_examples(self, filepath):
120
  """Yields examples."""