holylovenia commited on
Commit
94efbe5
1 Parent(s): 211d019

Add ASCEND loading script

Browse files
Files changed (1) hide show
  1. ascend.py +131 -0
ascend.py ADDED
@@ -0,0 +1,131 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # coding=utf-8
2
+ # Copyright 2021 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
+ """ Common Voice Dataset"""
16
+
17
+
18
+ import os
19
+
20
+ import datasets
21
+ from datasets.tasks import AutomaticSpeechRecognition
22
+ import pandas as pd
23
+
24
+
25
+ _CITATION = """\
26
+ @inproceedings{lovenia2021ascend,
27
+ title = {ASCEND: A Spontaneous Chinese-English Dataset for Code-switching in Multi-turn Conversation},
28
+ author = {Lovenia, Holy and Cahyawijaya, Samuel and Winata, Genta Indra and Xu, Peng and Yan, Xu and Liu, Zihan and Frieske, Rita and Yu, Tiezheng and Dai, Wenliang and Barezi, Elham J and others},
29
+ booktitle = {Proceedings of the International Conference on Language Resources and Evaluation, {LREC} 2022, 20-25 June 2022, Lu Palais du Pharo, France},
30
+ publisher = {European Language Resources Association},
31
+ year = {2022},
32
+ pages = {}
33
+ }
34
+ """
35
+
36
+ _DESCRIPTION = """\
37
+ ASCEND (A Spontaneous Chinese-English Dataset) introduces a high-quality resource of spontaneous multi-turn conversational dialogue Chinese-English code-switching corpus collected in Hong Kong. ASCEND consists of 10.62 hours of spontaneous speech with a total of ~12.3K utterances. The corpus is split into 3 sets: training, validation, and test with a ratio of 8:1:1 while maintaining a balanced gender proportion on each set.
38
+ """
39
+
40
+ _HOMEPAGE = "https://huggingface.co/datasets/CAiRE/ASCEND"
41
+
42
+
43
+ class ASCENDConfig(datasets.BuilderConfig):
44
+ """BuilderConfig for ASCEND."""
45
+
46
+ def __init__(self, name, **kwargs):
47
+ """
48
+ Args:
49
+ **kwargs: keyword arguments forwarded to super.
50
+ """
51
+ super(ASCENDConfig, self).__init__(name, **kwargs)
52
+
53
+
54
+ class ASCEND(datasets.GeneratorBasedBuilder):
55
+ """ASCEND: A Spontaneous Chinese-English Dataset for code-switching. Snapshot date: 5 January 2022."""
56
+
57
+ BUILDER_CONFIGS = [
58
+ ASCENDConfig(
59
+ name="train",
60
+ version=datasets.Version("1.0.0", ""),
61
+ description=_DESCRIPTION,
62
+ ),
63
+ ASCENDConfig(
64
+ name="validation",
65
+ version=datasets.Version("1.0.0", ""),
66
+ description=_DESCRIPTION,
67
+ ),
68
+ ASCENDConfig(
69
+ name="test",
70
+ version=datasets.Version("1.0.0", ""),
71
+ description=_DESCRIPTION,
72
+ ),
73
+ ]
74
+
75
+ def _info(self):
76
+ features = datasets.Features(
77
+ {
78
+ "path": datasets.Value("string"),
79
+ "audio": datasets.Audio(sampling_rate=16_000),
80
+ "transcription": datasets.Value("string"),
81
+ "duration": datasets.Value("float32"),
82
+ "language": datasets.Value("string"),
83
+ "original_speaker_id": datasets.Value("int64"),
84
+ "session_id": datasets.Value("int64"),
85
+ "topic": datasets.Value("string"),
86
+ }
87
+ )
88
+ return datasets.DatasetInfo(
89
+ description=_DESCRIPTION,
90
+ features=features,
91
+ supervised_keys=None,
92
+ homepage=_HOMEPAGE,
93
+ citation=_CITATION,
94
+ task_templates=[AutomaticSpeechRecognition(audio_column="audio", transcription_column="transcription")],
95
+ )
96
+
97
+ def _split_generators(self, dl_manager):
98
+ return [
99
+ datasets.SplitGenerator(
100
+ name=datasets.Split.TRAIN,
101
+ gen_kwargs={
102
+ "metadata_path": "train_metadata.csv",
103
+ "split": "train",
104
+ },
105
+ ),
106
+ datasets.SplitGenerator(
107
+ name=datasets.Split.TEST,
108
+ gen_kwargs={
109
+ "metadata_path": "test_metadata.csv",
110
+ "split": "test"
111
+ },
112
+ ),
113
+ datasets.SplitGenerator(
114
+ name=datasets.Split.VALIDATION,
115
+ gen_kwargs={
116
+ "metadata_path": "validation_metadata.csv",
117
+ "split": "validation",
118
+ },
119
+ ),
120
+ ]
121
+
122
+ def _generate_examples(self, metadata_path):
123
+ metadata_df = pd.read_csv(metadata_path)
124
+
125
+ for index, row in metadata_df.iterrows():
126
+ example = {
127
+ "id": index,
128
+ "audio": row["audio"],
129
+ "transcription": row["transcription"]
130
+ }
131
+ yield index, example