Source code for torchaudio.datasets.commonvoice
import os
from typing import List, Dict, Tuple
import torchaudio
from torchaudio.datasets.utils import download_url, extract_archive, unicode_csv_reader
from torch import Tensor
from torch.utils.data import Dataset
# Default TSV should be one of
# dev.tsv
# invalidated.tsv
# other.tsv
# test.tsv
# train.tsv
# validated.tsv
FOLDER_IN_ARCHIVE = "CommonVoice"
URL = "english"
VERSION = "cv-corpus-4-2019-12-10"
TSV = "train.tsv"
_CHECKSUMS = {
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/tt.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/en.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/de.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/fr.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/cy.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/br.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/cv.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/tr.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/ky.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/ga-IE.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/kab.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/ca.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/zh-TW.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/sl.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/it.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/nl.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/cnh.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/eo.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/et.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/fa.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/eu.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/es.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/zh-CN.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/mn.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/sah.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/dv.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/rw.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/sv-SE.tar.gz":
None,
"https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-3/ru.tar.gz":
None
}
def load_commonvoice_item(line: List[str],
header: List[str],
path: str,
folder_audio: str) -> Tuple[Tensor, int, Dict[str, str]]:
# Each line as the following data:
# client_id, path, sentence, up_votes, down_votes, age, gender, accent
assert header[1] == "path"
fileid = line[1]
filename = os.path.join(path, folder_audio, fileid)
waveform, sample_rate = torchaudio.load(filename)
dic = dict(zip(header, line))
return waveform, sample_rate, dic
[docs]class COMMONVOICE(Dataset):
"""Create a Dataset for CommonVoice.
Args:
root (str): Path to the directory where the dataset is found or downloaded.
tsv (str, optional): The name of the tsv file used to construct the metadata.
(default: ``"train.tsv"``)
url (str, optional): The URL to download the dataset from, or the language of
the dataset to download. (default: ``"english"``).
Allowed language values are ``"tatar"``, ``"english"``, ``"german"``,
``"french"``, ``"welsh"``, ``"breton"``, ``"chuvash"``, ``"turkish"``, ``"kyrgyz"``,
``"irish"``, ``"kabyle"``, ``"catalan"``, ``"taiwanese"``, ``"slovenian"``,
``"italian"``, ``"dutch"``, ``"hakha chin"``, ``"esperanto"``, ``"estonian"``,
``"persian"``, ``"portuguese"``, ``"basque"``, ``"spanish"``, ``"chinese"``,
``"mongolian"``, ``"sakha"``, ``"dhivehi"``, ``"kinyarwanda"``, ``"swedish"``,
``"russian"``, ``"indonesian"``, ``"arabic"``, ``"tamil"``, ``"interlingua"``,
``"latvian"``, ``"japanese"``, ``"votic"``, ``"abkhaz"``, ``"cantonese"`` and
``"romansh sursilvan"``.
folder_in_archive (str, optional): The top-level directory of the dataset.
version (str): Version string. (default: ``"cv-corpus-4-2019-12-10"``)
For the other allowed values, Please checkout https://commonvoice.mozilla.org/en/datasets.
download (bool, optional):
Whether to download the dataset if it is not found at root path. (default: ``False``).
"""
_ext_txt = ".txt"
_ext_audio = ".mp3"
_folder_audio = "clips"
def __init__(self,
root: str,
tsv: str = TSV,
url: str = URL,
folder_in_archive: str = FOLDER_IN_ARCHIVE,
version: str = VERSION,
download: bool = False) -> None:
languages = {
"tatar": "tt",
"english": "en",
"german": "de",
"french": "fr",
"welsh": "cy",
"breton": "br",
"chuvash": "cv",
"turkish": "tr",
"kyrgyz": "ky",
"irish": "ga-IE",
"kabyle": "kab",
"catalan": "ca",
"taiwanese": "zh-TW",
"slovenian": "sl",
"italian": "it",
"dutch": "nl",
"hakha chin": "cnh",
"esperanto": "eo",
"estonian": "et",
"persian": "fa",
"portuguese": "pt",
"basque": "eu",
"spanish": "es",
"chinese": "zh-CN",
"mongolian": "mn",
"sakha": "sah",
"dhivehi": "dv",
"kinyarwanda": "rw",
"swedish": "sv-SE",
"russian": "ru",
"indonesian": "id",
"arabic": "ar",
"tamil": "ta",
"interlingua": "ia",
"latvian": "lv",
"japanese": "ja",
"votic": "vot",
"abkhaz": "ab",
"cantonese": "zh-HK",
"romansh sursilvan": "rm-sursilv"
}
if url in languages:
ext_archive = ".tar.gz"
language = languages[url]
base_url = "https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com"
url = os.path.join(base_url, version, language + ext_archive)
basename = os.path.basename(url)
archive = os.path.join(root, basename)
basename = basename.rsplit(".", 2)[0]
folder_in_archive = os.path.join(folder_in_archive, version, basename)
self._path = os.path.join(root, folder_in_archive)
if download:
if not os.path.isdir(self._path):
if not os.path.isfile(archive):
checksum = _CHECKSUMS.get(url, None)
download_url(url, root, hash_value=checksum)
extract_archive(archive)
self._tsv = os.path.join(root, folder_in_archive, tsv)
with open(self._tsv, "r") as tsv:
walker = unicode_csv_reader(tsv, delimiter="\t")
self._header = next(walker)
self._walker = list(walker)
[docs] def __getitem__(self, n: int) -> Tuple[Tensor, int, Dict[str, str]]:
"""Load the n-th sample from the dataset.
Args:
n (int): The index of the sample to be loaded
Returns:
tuple: ``(waveform, sample_rate, dictionary)``, where dictionary is built
from the TSV file with the following keys: ``client_id``, ``path``, ``sentence``,
``up_votes``, ``down_votes``, ``age``, ``gender`` and ``accent``.
"""
line = self._walker[n]
return load_commonvoice_item(line, self._header, self._path, self._folder_audio)
def __len__(self) -> int:
return len(self._walker)
