|
import os |
|
import datasets |
|
|
|
from pathlib import Path |
|
|
|
from tqdm import tqdm |
|
|
|
from datasets.tasks import AutomaticSpeechRecognition |
|
|
|
logger = datasets.utils.logging.get_logger(__name__) |
|
|
|
|
|
_DESCRIPTION = """ |
|
Dyma fanc o 25 awr 34 munud a 24 eiliad o segmentau o leferydd naturiol dros hanner cant o gyfranwyr ar ffurf ffeiliau mp3, ynghyd â thrawsgrifiadau 'verbatim' cyfatebol o’r lleferydd ar ffurf ffeil .tsv. Mae'r mwyafrif o'r lleferydd yn leferydd digymell, naturiol. Dosbarthwn y deunydd hwn o dan drwydded agored CC0. |
|
|
|
This resource is a bank of 25 hours 34 minutes and 24 seconds of segments of natural speech from over 50 contributors in mp3 file format, together with corresponding 'verbatim' transcripts of the speech in .tsv file format. The majority of the speech is spontaneous, natural speech. We distribute this material under a CC0 open license. |
|
""" |
|
|
|
|
|
_CITATION = """ |
|
} |
|
""" |
|
|
|
_HOMEPAGE = "https://git.techiaith.bangor.ac.uk/data-porth-technolegau-iaith/banc-trawsgrifiadau-bangor" |
|
|
|
_LICENSE = "Creative Commons Zero v1.0 Universal" |
|
|
|
_DL_URL = "https://huggingface.co/datasets/DewiBrynJones/banc-trawsgrifiadau-bangor/resolve/main" |
|
|
|
|
|
|
|
|
|
|
|
class BancTrawsgrifiadauBangor(datasets.GeneratorBasedBuilder): |
|
|
|
VERSION = datasets.Version("23.6.0","") |
|
|
|
def _info(self): |
|
|
|
features=datasets.Features( |
|
{ |
|
"path": datasets.Value("string"), |
|
"audio": datasets.Audio(sampling_rate=48_000), |
|
"sentence": datasets.Value("string"), |
|
} |
|
) |
|
|
|
|
|
return datasets.DatasetInfo( |
|
description=_DESCRIPTION, |
|
features=features, |
|
homepage=_HOMEPAGE, |
|
license=_LICENSE, |
|
citation=_CITATION, |
|
task_templates=[AutomaticSpeechRecognition(audio_column="audio", transcription_column="sentence")] |
|
) |
|
|
|
|
|
def _split_generators(self, dl_manager): |
|
|
|
clips_data_dir = os.path.join(dl_manager.download_and_extract(os.path.join(_DL_URL, "clips.zip")), "clips") |
|
|
|
return [ |
|
datasets.SplitGenerator( |
|
name="clips", |
|
gen_kwargs={ |
|
"metadata_filepath": dl_manager.download(f"{_DL_URL}/clips.tsv"), |
|
"path_to_clips": clips_data_dir, |
|
}, |
|
), |
|
datasets.SplitGenerator( |
|
name=datasets.Split.TRAIN, |
|
gen_kwargs={ |
|
"metadata_filepath": dl_manager.download(f"{_DL_URL}/train.tsv"), |
|
"path_to_clips": clips_data_dir, |
|
}, |
|
), |
|
datasets.SplitGenerator( |
|
name=datasets.Split.TEST, |
|
gen_kwargs={ |
|
"metadata_filepath": dl_manager.download(f"{_DL_URL}/test.tsv"), |
|
"path_to_clips": clips_data_dir, |
|
}, |
|
), |
|
] |
|
|
|
|
|
def _generate_examples(self, metadata_filepath, path_to_clips): |
|
|
|
|
|
|
|
with open(metadata_filepath, encoding='utf-8') as f: |
|
lines = f.readlines() |
|
|
|
for id_, line in tqdm(enumerate(lines[1:]), total=len(lines)) : |
|
field_values = line.strip().split("\t") |
|
|
|
|
|
|
|
audio_file_path=os.path.join(path_to_clips, field_values[0]) |
|
transcript = field_values[2] |
|
|
|
|
|
row = { "path": audio_file_path, "sentence": transcript } |
|
row["audio"] = {"audio_path": audio_file_path, "bytes": open(audio_file_path,'rb').read()} |
|
|
|
yield id_, row |
|
|
