| import os
|
| import csv
|
| import datasets
|
| from tqdm import tqdm
|
|
|
| _DESCRIPTION = "A speech dataset designed for automatic speech recognition (ASR), structured like Mozilla Common Voice."
|
| _CITATION = "No citation available yet."
|
|
|
| class MyDatasetConfig(datasets.BuilderConfig):
|
| def __init__(self, **kwargs):
|
| super(MyDatasetConfig, self).__init__(**kwargs)
|
|
|
| class MyDataset(datasets.GeneratorBasedBuilder):
|
| DEFAULT_WRITER_BATCH_SIZE = 1000
|
|
|
| BUILDER_CONFIGS = [
|
| MyDatasetConfig(
|
| name="default",
|
| version=datasets.Version("1.0.0"),
|
| description=_DESCRIPTION,
|
| ),
|
| ]
|
|
|
| def _info(self):
|
| features = datasets.Features({
|
| "client_id": datasets.Value("string"),
|
| "path": datasets.Value("string"),
|
| "audio": datasets.features.Audio(sampling_rate=16000),
|
| "text": datasets.Value("string"),
|
| "up_votes": datasets.Value("int64"),
|
| "down_votes": datasets.Value("int64"),
|
| "age": datasets.Value("string"),
|
| "gender": datasets.Value("string"),
|
| "accent": datasets.Value("string"),
|
| "locale": datasets.Value("string"),
|
| "segment": datasets.Value("string"),
|
| "variant": datasets.Value("string"),
|
| })
|
|
|
| return datasets.DatasetInfo(
|
| description=_DESCRIPTION,
|
| features=features,
|
| supervised_keys=None,
|
| citation=_CITATION,
|
| version=self.config.version,
|
| )
|
|
|
| def _split_generators(self, dl_manager):
|
| data_dir = self.config.data_dir
|
|
|
| splits = ["train", "validation", "test"]
|
| split_generators = []
|
|
|
| for split in splits:
|
| audio_tar = os.path.join(data_dir, f"{split}_audio")
|
| metadata_csv = os.path.join(data_dir, f"{split}_metadata.csv")
|
|
|
| if os.path.exists(audio_tar) and os.path.exists(metadata_csv):
|
| split_generators.append(
|
| datasets.SplitGenerator(
|
| name=getattr(datasets.Split, split.upper()),
|
| gen_kwargs={
|
| "archives": dl_manager.iter_archive(audio_tar),
|
| "metadata_path": metadata_csv,
|
| },
|
| )
|
| )
|
|
|
| return split_generators
|
|
|
| def _generate_examples(self, archives, metadata_path):
|
|
|
| metadata = {}
|
| data_fields = list(self._info().features.keys())
|
|
|
| with open(metadata_path, encoding="utf-8-sig") as f:
|
| reader = csv.DictReader(f, delimiter=",", quoting=csv.QUOTE_NONE)
|
| reader.fieldnames = [name.strip().replace('"', '') for name in reader.fieldnames]
|
|
|
| for row in tqdm(reader, desc="Loading metadata..."):
|
| row = {k.replace('"', ''): v.replace('"', '') for k, v in row.items()}
|
| if not row["file_name"].endswith(".wav"):
|
| row["file_name"] += ".wav"
|
|
|
|
|
| if "accents" in row:
|
| row["accent"] = row["accents"]
|
| del row["accents"]
|
|
|
|
|
| for field in data_fields:
|
| if field not in row:
|
|
|
| if field in ["up_votes", "down_votes"]:
|
| row[field] = 0
|
| else:
|
| row[field] = ""
|
|
|
| metadata[row["file_name"]] = row
|
|
|
|
|
|
|
| for i, audio_archive in enumerate(archives):
|
| for path_in_tar, file_obj in audio_archive:
|
| _, filename = os.path.split(path_in_tar)
|
|
|
| if filename in metadata:
|
| example = dict(metadata[filename])
|
|
|
|
|
| example["audio"] = {"path": path_in_tar, "bytes": file_obj.read()}
|
| example["path"] = path_in_tar
|
|
|
| yield path_in_tar, example
|
|
|
