File size: 6,362 Bytes
2bdfded
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8761f7d
 
 
 
 
 
 
 
 
 
2bdfded
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
""" Kathbath Dataset"""

import csv
import os

import datasets
from datasets.utils.py_utils import size_str

from .languages import LANGUAGES
from .release_stats import STATS

_CITATION = """\
@misc{https://doi.org/10.48550/arxiv.2208.11761,
  doi = {10.48550/ARXIV.2208.11761},
  url = {https://arxiv.org/abs/2208.11761},
  author = {Javed, Tahir and Bhogale, Kaushal Santosh and Raman, Abhigyan and Kunchukuttan, Anoop and Kumar, Pratyush and Khapra, Mitesh M.},
  title = {IndicSUPERB: A Speech Processing Universal Performance Benchmark for Indian languages},
  publisher = {arXiv},
  year = {2022},
  copyright = {arXiv.org perpetual, non-exclusive license}
}
"""

_HOMEPAGE = "https://ai4bharat.iitm.ac.in/indic-superb/"

_LICENSE = "https://creativecommons.org/publicdomain/zero/1.0/"

_DATA_URL = "https://huggingface.co/datasets/ai4bharat/kathbath/resolve/main/data"


class KathbathConfig(datasets.BuilderConfig):
    """BuilderConfig for Kathbath."""

    def __init__(self, name, version, **kwargs):
        self.language = kwargs.pop("language", None)
        self.release_date = kwargs.pop("release_date", None)
        self.num_clips = kwargs.pop("num_clips", None)
        self.num_speakers = kwargs.pop("num_speakers", None)
        self.total_hr = kwargs.pop("total_hr", None)
        self.size_bytes = kwargs.pop("size_bytes", None)
        self.size_human = size_str(self.size_bytes)
        description = (
            f"Kathbath speech to text dataset in {self.language} released on {self.release_date}. "
            f"The dataset comprises {self.total_hr} hours of transcribed speech data"
        )
        super(KathbathConfig, self).__init__(
            name=name,
            version=datasets.Version(version),
            description=description,
            **kwargs,
        )


class Kathbath(datasets.GeneratorBasedBuilder):
    DEFAULT_CONFIG_NAME = "_all_"

    BUILDER_CONFIGS = [
        KathbathConfig(
            name=lang,
            version=STATS["version"],
            language=LANGUAGES[lang],
            release_date=STATS["date"],
            # num_clips=lang_stats["clips"],
            # num_speakers=lang_stats["users"],
            total_hr=float(lang_stats["totalHrs"]) if lang_stats["totalHrs"] else None,
            size_bytes=int(lang_stats["size"]) if lang_stats["size"] else None,
        )
        for lang, lang_stats in STATS["locales"].items()
    ]

    def _info(self):
        total_languages = len(STATS["locales"])
        total_hours = self.config.total_hr
        description = (
            "LibriVox-Indonesia is a speech dataset generated from LibriVox with only languages from Indonesia."
            f"The dataset currently consists of {total_hours} hours of speech "
            f"in {total_languages} languages, but more voices and languages are always added."
        )
        features = datasets.Features(
            {
                "path": datasets.Value("string"),
                "language": datasets.Value("string"),
                "speaker": datasets.Value("string"),
                "sentence": datasets.Value("string"),
                "audio": datasets.features.Audio(sampling_rate=16000)
            }
        )

        return datasets.DatasetInfo(
            description=description,
            features=features,
            supervised_keys=None,
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
            version=self.config.version,
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        dl_manager.download_config.ignore_url_params = True
        audio_path = {}
        local_extracted_archive = {}
        metadata_path = {}
        split_type = {"train": datasets.Split.TRAIN, "valid": datasets.Split.VALIDATION, "test_unknown": datasets.Split.TEST, "test_known": datasets.Split.TEST}
        for split in split_type:
            
            if split == 'train':
                audio_paths = [
                    f"{_DATA_URL}/audio_{split}.tar.partaa",
                    f"{_DATA_URL}/audio_{split}.tar.partab",
                    f"{_DATA_URL}/audio_{split}.tar.partac",
                ]
            else:
                audio_paths = [f"{_DATA_URL}/audio_{split}.tar"]
            audio_path[split] = dl_manager.download(audio_paths)
            local_extracted_archive[split] = dl_manager.extract(audio_path[split]) if not dl_manager.is_streaming else None
            metadata_path[split] = dl_manager.download(f"{_DATA_URL}/metadata_{split}.tsv")
        path_to_clips = "kb_data_clean_m4a"

        return [
            datasets.SplitGenerator(
                name=split_type[split],
                gen_kwargs={
                    "local_extracted_archive": local_extracted_archive[split],
                    "audio_files": dl_manager.iter_archive(audio_path[split]),
                    "metadata_path": metadata_path[split],
                    "path_to_clips": path_to_clips,
                },
            ) for split in split_type
        ]

    def _generate_examples(
        self,
        local_extracted_archive,
        audio_files,
        metadata_path,
        path_to_clips,
    ):
        """Yields examples."""
        data_fields = list(self._info().features.keys())
        metadata = {}
        with open(metadata_path, "r", encoding="utf-8") as f:
            reader = csv.DictReader(f, delimiter="\t")
            for row in reader:
                if self.config.name == "_all_" or self.config.name == row["language"]:
                    row["path"] = os.path.join(path_to_clips, row["path"])
                    # if data is incomplete, fill with empty values
                    for field in data_fields:
                        if field not in row:
                            row[field] = ""
                    metadata[row["path"]] = row
        id_ = 0
        for path, f in audio_files:
            if path in metadata:
                result = dict(metadata[path])
                # set the audio feature and the path to the extracted file
                path = os.path.join(local_extracted_archive, path) if local_extracted_archive else path
                result["audio"] = {"path": path, "bytes": f.read()}
                result["path"] = path
                yield id_, result
                id_ += 1