secondamericasnlp2022.py

_LANGUAGES = {
    "qu": "quechua",
    "gn": "guarani",
}
_VERSION = "1.0.0"


def _info(self):
    return datasets.DatasetInfo(
        description=_DESCRIPTION,
        features=datasets.Features(
            {
                "audio": datasets.Value("string"),
                "source_processed": datasets.Value("string"),
                "source_raw": datasets.Value("string"),
                "target_raw": datasets.Value("string"),
                "split": datasets.Value("string")
                "subset": datasets.Value("string")
            }
        ),
        supervised_keys = None,
    )

class SecondAmericasNLP2022Config(datasets.BuilderConfig):
    def __init__(self, languages=None, **kwargs):
        super(SecondAmericasNLP2022Config, self).__init__(version=datasets.Version(_VERSION, ""), **kwargs),
        self.languages = languages

class SecondAmericasNLP2022(datasets.GeneratorBasedBuilder):

    BUILDER_CONFIG_CLASS = SecondAmericasNLP2022Config

    BUILDER_CONFIGS = [
        SecondAmericasNLP2022Config(
            name=_ALL_LANGUAGES,
            languages=_LANGUAGES,
        )
    ] + [
        TapacoConfig(
            name=lang,
            languages=[lang],
        )
        for lang in _LANGUAGES
    ]
    BUILDER_CONFIG_CLASS = SecondAmericasNLP2022Config
    DEFAULT_CONFIG_NAME = _ALL_LANGUAGES