holylovenia commited on
Commit
e161039
·
verified ·
1 Parent(s): 1a71cfd

Upload wikihow_gosc.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. wikihow_gosc.py +151 -0
wikihow_gosc.py ADDED
@@ -0,0 +1,151 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2022 The HuggingFace Datasets Authors and the current dataset script contributor.
3
+ #
4
+ # Licensed under the Apache License, Version 2.0 (the "License");
5
+ # you may not use this file except in compliance with the License.
6
+ # You may obtain a copy of the License at
7
+ #
8
+ # http://www.apache.org/licenses/LICENSE-2.0
9
+ #
10
+ # Unless required by applicable law or agreed to in writing, software
11
+ # distributed under the License is distributed on an "AS IS" BASIS,
12
+ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13
+ # See the License for the specific language governing permissions and
14
+ # limitations under the License.
15
+ import json
16
+ import os
17
+ from pathlib import Path
18
+ from typing import Dict, List, Tuple
19
+
20
+ import datasets
21
+
22
+ from seacrowd.utils.configs import SEACrowdConfig
23
+ from seacrowd.utils.constants import Licenses, Tasks
24
+
25
+ _CITATION = """
26
+ @inproceedings{lyu-etal-2021-goal,
27
+ title = "Goal-Oriented Script Construction",
28
+ author = "Lyu, Qing and
29
+ Zhang, Li and
30
+ Callison-Burch, Chris",
31
+ editor = "Belz, Anya and
32
+ Fan, Angela and
33
+ Reiter, Ehud and
34
+ Sripada, Yaji",
35
+ booktitle = "Proceedings of the 14th International Conference on Natural Language Generation",
36
+ month = aug,
37
+ year = "2021",
38
+ publisher = "Association for Computational Linguistics",
39
+ url = "https://aclanthology.org/2021.inlg-1.19",
40
+ doi = "10.18653/v1/2021.inlg-1.19",
41
+ pages = "184--200",
42
+ }
43
+ """
44
+ _LOCAL = False
45
+ _LANGUAGES = {"ind": "id", "tha": "th", "vie": "vn"}
46
+ _DATASETNAME = "wikihow_gosc"
47
+ _DESCRIPTION = """
48
+ This dataset consists of wikiHow goal-oriented scripts. For each goal or task, sections with steps to achieve this task are
49
+ generated. Both the sections and steps within them are classified as either ordered or unordered.
50
+ """
51
+
52
+ _HOMEPAGE = "https://github.com/veronica320/wikihow-GOSC/tree/main?tab=readme-ov-file"
53
+ _LICENSE = Licenses.MIT.value
54
+ _URL = "https://drive.google.com/uc?id=1AqAocrNFEPhBAfa5ATCj-3xMWbq659ME"
55
+
56
+ _SUPPORTED_TASKS = [Tasks.INSTRUCTION_TUNING]
57
+ _SOURCE_VERSION = "1.0.0"
58
+ _SEACROWD_VERSION = "2024.06.20"
59
+
60
+
61
+ class WikiHowGOSCDataset(datasets.GeneratorBasedBuilder):
62
+ """Dataset of WikiHow tasks/goals with generated steps to perform them."""
63
+
64
+ SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
65
+
66
+ BUILDER_CONFIGS = [
67
+ SEACrowdConfig(
68
+ name=f"{_DATASETNAME}_{lang}_source",
69
+ version=_SOURCE_VERSION,
70
+ description=f"{_DATASETNAME} source schema for {lang} language",
71
+ schema="source",
72
+ subset_id=f"{_DATASETNAME}_{lang}",
73
+ )
74
+ for lang in _LANGUAGES
75
+ ]
76
+
77
+ DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_ind_source"
78
+
79
+ def _info(self) -> datasets.DatasetInfo:
80
+
81
+ features = datasets.Features(
82
+ {
83
+ "title": datasets.Value("string"),
84
+ "category": datasets.Value("string"),
85
+ "sections": datasets.Sequence({"section": datasets.Value("string"), "steps": datasets.Sequence(datasets.Value("string")), "ordered": datasets.Value("int32")}),
86
+ "ordered": datasets.Value("int32"),
87
+ }
88
+ )
89
+
90
+ return datasets.DatasetInfo(
91
+ description=_DESCRIPTION,
92
+ features=features,
93
+ homepage=_HOMEPAGE,
94
+ license=_LICENSE,
95
+ citation=_CITATION,
96
+ )
97
+
98
+ def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
99
+ """Returns SplitGenerators."""
100
+ try:
101
+ import gdown
102
+ except ImportError:
103
+ raise ImportError("Please install `gdown` to enable downloading data from google drive.")
104
+
105
+ # Download from Google drive
106
+ output_dir = Path.cwd() / "data" / "wikihow_gosc"
107
+ output_dir.mkdir(parents=True, exist_ok=True)
108
+ output_file = output_dir / "wikihow_multilingual_scripts.zip"
109
+ if not output_file.exists():
110
+ gdown.download(_URL, str(output_file), fuzzy=True)
111
+ else:
112
+ print(f"File already downloaded: {str(output_file)}")
113
+
114
+ data_dir = Path(dl_manager.extract(output_file))
115
+ lang = _LANGUAGES[self.config.subset_id.split("_")[-1]]
116
+
117
+ return [ # Train and test are in same file
118
+ datasets.SplitGenerator(
119
+ name=datasets.Split.TRAIN,
120
+ gen_kwargs={
121
+ "filepath": os.path.join(data_dir, f"script_{lang}.json"),
122
+ "split": "train",
123
+ },
124
+ ),
125
+ datasets.SplitGenerator(
126
+ name=datasets.Split.TEST,
127
+ gen_kwargs={
128
+ "filepath": os.path.join(data_dir, f"script_{lang}.json"),
129
+ "split": "test",
130
+ },
131
+ ),
132
+ ]
133
+
134
+ def _generate_examples(self, filepath: Path, split: str) -> Tuple[int, Dict]:
135
+ """Yields examples as (key, example) tuples."""
136
+ with open(filepath, "r", encoding="utf-8") as file:
137
+ data = json.load(file)
138
+ for key, example in enumerate(data[split]):
139
+ if "sections" not in example: # Single-section example
140
+ yield key, {
141
+ "title": example["title"],
142
+ "category": example["category"],
143
+ "sections": [{
144
+ "section": "",
145
+ "steps": example["steps"],
146
+ "ordered": example["ordered"],
147
+ }],
148
+ "ordered": 1
149
+ }
150
+ else:
151
+ yield key, example