hpprc commited on
Commit
67b824d
1 Parent(s): b8ee708

:sparkles: Add the builder scritp

Browse files
Files changed (3) hide show
  1. janli.py +75 -0
  2. poetry.lock +0 -0
  3. pyproject.toml +23 -0
janli.py ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import pandas as pd
2
+
3
+ import datasets as ds
4
+
5
+ _CITATION = """\
6
+ @InProceedings{yanaka-EtAl:2021:blackbox,
7
+ author = {Yanaka, Hitomi and Mineshima, Koji},
8
+ title = {Assessing the Generalization Capacity of Pre-trained Language Models through Japanese Adversarial Natural Language Inference},
9
+ booktitle = {Proceedings of the 2021 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP (BlackboxNLP2021)},
10
+ year = {2021},
11
+ }
12
+ """
13
+
14
+ _DESCRIPTION = """\
15
+
16
+ """
17
+
18
+ _HOMEPAGE = "https://github.com/verypluming/JaNLI"
19
+
20
+ _LICENSE = "CC BY-SA 4.0"
21
+
22
+ _DOWNLOAD_URL = "https://raw.githubusercontent.com/verypluming/JaNLI/main/janli.tsv"
23
+
24
+
25
+ class JaNLIDataset(ds.GeneratorBasedBuilder):
26
+ VERSION = ds.Version("1.0.0")
27
+
28
+ BUILDER_CONFIGS = [
29
+ ds.BuilderConfig(
30
+ name="default",
31
+ version=VERSION,
32
+ description=_DESCRIPTION,
33
+ ),
34
+ ]
35
+
36
+ def _info(self) -> ds.DatasetInfo:
37
+ features = ds.Features(
38
+ {
39
+ "id": ds.Value("int64"),
40
+ "sentence_A_Ja": ds.Value("string"),
41
+ "sentence_B_Ja": ds.Value("string"),
42
+ "entailment_label_Ja": ds.ClassLabel(names=["entailment", "non-entailment"]),
43
+ "heuristics": ds.Value("string"),
44
+ "number_of_NPs": ds.Value("int32"),
45
+ "semtag": ds.Value("string"),
46
+ }
47
+ )
48
+ return ds.DatasetInfo(
49
+ description=_DESCRIPTION,
50
+ citation=_CITATION,
51
+ homepage=_HOMEPAGE,
52
+ license=_LICENSE,
53
+ features=features,
54
+ )
55
+
56
+ def _split_generators(self, dl_manager: ds.DownloadManager):
57
+ data_path = dl_manager.download_and_extract(_DOWNLOAD_URL)
58
+ df: pd.DataFrame = pd.read_table(data_path, header=0, sep="\t", index_col=0)
59
+ df["id"] = df.index
60
+
61
+ return [
62
+ ds.SplitGenerator(
63
+ name=ds.Split.TRAIN,
64
+ gen_kwargs={"df": df[df["split"] == "train"]},
65
+ ),
66
+ ds.SplitGenerator(
67
+ name=ds.Split.TEST,
68
+ gen_kwargs={"df": df[df["split"] == "test"]},
69
+ ),
70
+ ]
71
+
72
+ def _generate_examples(self, df: pd.DataFrame):
73
+ df = df.drop("split", axis=1)
74
+ for i, row in enumerate(df.to_dict("records")):
75
+ yield i, row
poetry.lock ADDED
The diff for this file is too large to render. See raw diff
 
pyproject.toml ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [tool.poetry]
2
+ name = "datasets-janli"
3
+ version = "0.1.0"
4
+ description = ""
5
+ authors = ["hppRC <hpp.ricecake@gmail.com>"]
6
+ readme = "README.md"
7
+ packages = []
8
+
9
+ [tool.poetry.dependencies]
10
+ python = "^3.8.1"
11
+ datasets = "^2.11.0"
12
+
13
+
14
+ [tool.poetry.group.dev.dependencies]
15
+ black = "^22.12.0"
16
+ isort = "^5.11.4"
17
+ flake8 = "^6.0.0"
18
+ mypy = "^0.991"
19
+ pytest = "^7.2.0"
20
+
21
+ [build-system]
22
+ requires = ["poetry-core"]
23
+ build-backend = "poetry.core.masonry.api"