shunk031 commited on
Commit
4f56210
·
unverified ·
1 Parent(s): e8929fb

Update readme and dataset name (#2)

Browse files

* update script

* update README

Files changed (3) hide show
  1. README.md +30 -0
  2. jsnli.py +4 -5
  3. tests/jsnli_test.py +2 -2
README.md CHANGED
@@ -1,3 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  # Dataset Card for JSNLI
2
 
3
  [![CI](https://github.com/shunk031/huggingface-datasets_jsnli/actions/workflows/ci.yaml/badge.svg)](https://github.com/shunk031/huggingface-datasets_jsnli/actions/workflows/ci.yaml)
 
1
+ ---
2
+ language:
3
+ - ja
4
+
5
+ license:
6
+ - cc-by-sa-4.0
7
+
8
+ multilinguality:
9
+ - monolingual
10
+
11
+ task_categories:
12
+ - text-classification
13
+
14
+ task_ids:
15
+ - natural-language-inference
16
+ - multi-input-text-classification
17
+
18
+ tags:
19
+ - natural-language-inference
20
+ - nli
21
+ - jsnli
22
+
23
+ datasets:
24
+ - without-filtering
25
+ - with-filtering
26
+
27
+ metrics:
28
+ - accuracy
29
+ ---
30
+
31
  # Dataset Card for JSNLI
32
 
33
  [![CI](https://github.com/shunk031/huggingface-datasets_jsnli/actions/workflows/ci.yaml/badge.svg)](https://github.com/shunk031/huggingface-datasets_jsnli/actions/workflows/ci.yaml)
jsnli.py CHANGED
@@ -30,15 +30,14 @@ _URL = "https://nlp.ist.i.kyoto-u.ac.jp/DLcounter/lime.cgi?down=https://nlp.ist.
30
 
31
  class JSNLIDataset(ds.GeneratorBasedBuilder):
32
  VERSION = ds.Version("1.1.0") # type: ignore
33
- DEFAULT_CONFIG_NAME: str = "w_filtering" # type: ignore
34
  BUILDER_CONFIG = [
35
  ds.BuilderConfig(
36
- name="w_filtering",
37
  version=VERSION, # type: ignore
38
  description="SNLIの学習データに機械翻訳を適用した後、BLEUスコアの閾値0.1でフィルタリングを施したもの。BERTにこの学習データを学習させることにより、93.0%の精度を記録した。(533,005ペア)",
39
  ),
40
  ds.BuilderConfig(
41
- name="wo_filtering",
42
  version=VERSION, # type: ignore
43
  description="SNLIの学習データに機械翻訳を適用したもの。フィルタリングは行っていない。(548,014ペア)",
44
  ),
@@ -70,9 +69,9 @@ class JSNLIDataset(ds.GeneratorBasedBuilder):
70
  train_wo_filtering_path = os.path.join(jsnli_dir, "train_wo_filtering.tsv")
71
 
72
  dev_path = os.path.join(jsnli_dir, "dev.tsv")
73
- if "w_filtering" in self.config.name:
74
  tng_path = train_w_filtering_path
75
- elif "wo_filtering" in self.config.name:
76
  tng_path = train_wo_filtering_path
77
  else:
78
  raise ValueError(f"Invalid config name: {self.config.name}")
 
30
 
31
  class JSNLIDataset(ds.GeneratorBasedBuilder):
32
  VERSION = ds.Version("1.1.0") # type: ignore
 
33
  BUILDER_CONFIG = [
34
  ds.BuilderConfig(
35
+ name="with-filtering",
36
  version=VERSION, # type: ignore
37
  description="SNLIの学習データに機械翻訳を適用した後、BLEUスコアの閾値0.1でフィルタリングを施したもの。BERTにこの学習データを学習させることにより、93.0%の精度を記録した。(533,005ペア)",
38
  ),
39
  ds.BuilderConfig(
40
+ name="without-filtering",
41
  version=VERSION, # type: ignore
42
  description="SNLIの学習データに機械翻訳を適用したもの。フィルタリングは行っていない。(548,014ペア)",
43
  ),
 
69
  train_wo_filtering_path = os.path.join(jsnli_dir, "train_wo_filtering.tsv")
70
 
71
  dev_path = os.path.join(jsnli_dir, "dev.tsv")
72
+ if "with-filtering" in self.config.name:
73
  tng_path = train_w_filtering_path
74
+ elif "without-filtering" in self.config.name:
75
  tng_path = train_wo_filtering_path
76
  else:
77
  raise ValueError(f"Invalid config name: {self.config.name}")
tests/jsnli_test.py CHANGED
@@ -10,8 +10,8 @@ def dataset_path() -> str:
10
  @pytest.mark.parametrize(
11
  "dataset_name, expected_num_train,",
12
  (
13
- ("wo_filtering", 548014),
14
- ("w_filtering", 533005),
15
  ),
16
  )
17
  def test_load_dataset(
 
10
  @pytest.mark.parametrize(
11
  "dataset_name, expected_num_train,",
12
  (
13
+ ("without-filtering", 548014),
14
+ ("with-filtering", 533005),
15
  ),
16
  )
17
  def test_load_dataset(