voxpopuli.py

from collections import defaultdict
import os
import glob
import csv
from tqdm.auto import tqdm

import datasets


_LANGUAGES = sorted(
    [
        "en", "de", "fr", "es", "pl", "it", "ro", "hu", "cs", "nl", "fi", "hr",
        "sk", "sl", "et", "lt", "pt", "bg", "el", "lv", "mt", "sv", "da"
    ]
)
_LANGUAGES_V2 = [f"{x}_v2" for x in _LANGUAGES]

_YEARS = list(range(2009, 2020 + 1))

# unnecessary
_CONFIG_TO_LANGS = {
    "400k": _LANGUAGES,
    "100k": _LANGUAGES,
    "10k": _LANGUAGES,
}

_CONFIG_TO_YEARS = {
    "400k": _YEARS + [f"{y}_2" for y in _YEARS],
    "100k": _YEARS,
    "10k": [2019, 2020],
    # "asr": _YEARS
}
for lang in _LANGUAGES:
    _CONFIG_TO_YEARS[lang] = _YEARS

_BASE_URL = "https://dl.fbaipublicfiles.com/voxpopuli/"

_DATA_URL = _BASE_URL + "audios/{lang}_{year}.tar"

_META_URL = _BASE_URL + "annotations/unlabelled_v2.tsv.gz"


class VoxpopuliConfig(datasets.BuilderConfig):
    """BuilderConfig for VoxPopuli."""

    def __init__(self, name, **kwargs):
        """
        Args:
          name: `string`, name of dataset config
          **kwargs: keyword arguments forwarded to super.
        """
        super().__init__(name=name, **kwargs)
        self.languages = [name] if name in _LANGUAGES else _LANGUAGES


class Voxpopuli(datasets.GeneratorBasedBuilder):
    """The Voxpopuli dataset."""

    VERSION = datasets.Version("1.0.0")  # TODO ??
    BUILDER_CONFIGS = [
        VoxpopuliConfig(
            name=name,
            # version=VERSION,
            description="",  # TODO
            )
        for name in _LANGUAGES + ["10k", "100k", "400k"]
    ]
    # DEFAULT_CONFIG_NAME = "400k"
    # DEFAULT_WRITER_BATCH_SIZE = 256

    def _info(self):
        features = datasets.Features(
            {
                "path": datasets.Value("string"),
                "language": datasets.ClassLabel(names=_LANGUAGES),
                "year": datasets.Value("int16"),
                "audio": datasets.Audio(sampling_rate=16_000),
                "segment_id": datasets.Value("int16"),
            }
        )
        return datasets.DatasetInfo(
            # description=_DESCRIPTION,
            features=features,
            # homepage=_HOMEPAGE,
            # license=_LICENSE,
            # citation=_CITATION,
        )

    def _read_metadata(self, metadata_path):
        # TODO: check for predicate??
        #  @ https://github.com/facebookresearch/voxpopuli/blob/main/voxpopuli/get_unlabelled_data.py#L34
        metadata = defaultdict(list)

        with open(metadata_path, encoding="utf-8") as csv_file:
            csv_reader = csv.reader(csv_file, delimiter="\t")
            for i, row in tqdm(enumerate(csv_reader)):
                if i == 0:
                    continue
                audio_id, segment_id, start, end = row
                event_id, lang = audio_id.rsplit("_", 1)[-2:]
                if lang in self.languages:
                # if lang in ["hr", "et"]:
                    metadata[audio_id].append((float(start), float(end)))

        return metadata

    def _split_generators(self, dl_manager):
        metadata_path = dl_manager.download_and_extract(_META_URL)

        years = _CONFIG_TO_YEARS[self.config.name]
        # urls = [_DATA_URL.format(lang=language, year=year) for language in ["hr", "et"] for year in [2020]]  # , "et"]
        urls = [_DATA_URL.format(lang=language, year=year) for language in self.languages for year in years]
        dl_manager.download_config.num_proc = len(urls)
        data_dirs = dl_manager.download_and_extract(urls)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "data_dirs": data_dirs,
                    "metadata_path": metadata_path,
                }
            ),
        ]

    def _generate_examples(self, data_dirs, metadata_path):
        try:
            import torch
            import torchaudio
        except ImportError as e:
            raise ValueError(
                "Loading voxpopuli requires `torchaudio` to be installed."
                "You can install torchaudio with `pip install torchaudio`." + e
            )

        metadata = self._read_metadata(metadata_path)

        for data_dir in data_dirs:
            for file in glob.glob(f"{data_dir}/**/*.ogg", recursive=True):
                path_components = file.split(os.sep)
                language, year, audio_filename = path_components[-3:]
                audio_id, _ = os.path.splitext(audio_filename)
                timestamps = metadata[audio_id]

                waveform, sr = torchaudio.load(file)
                duration = waveform.size(1)

                # split audio on the fly and write segments as arrays
                for segment_id, (start, stop) in enumerate(timestamps):
                    segment = waveform[:, int(start * sr): min(int(stop * sr), duration)]

                    yield f"{audio_filename}_{segment_id}", {
                        "path": file,
                        "language": language,
                        "year": year,
                        "audio": {
                            "array": segment[0],  # segment is a 2-dim array
                            "sampling_rate": 16_000
                        },
                        "segment_id": segment_id,
                    }