bertic_data.py

import datasets
import gzip
import json


_URL = "http://nl.ijs.si/nikola/dedup_hbs/"
_URLS = [
    # "macocu.hbs.translit.dedup.lines.gz",
    # "hr_news.translit.dedup.lines.gz",
    # "srwac.translit.dedup.lines.gz",
    "riznica.translit.dedup.lines.gz",
    # "mC4.sr.translit.dedup.lines.gz",
    # "hrwac.translit.dedup.lines.gz",
    # "cnrwac.translit.dedup.lines.gz",
    # "classla-sr.translit.dedup.lines.gz",
    # "classla-hr.translit.dedup.lines.gz",
    #  "classla-bs.translit.dedup.lines.gz",
    #  "cc100-sr.translit.dedup.lines.gz",
    #  "cc100-hr.translit.dedup.lines.gz",
    #  "bswac.translit.dedup.lines.gz",
]
_URLS = [_URL + i for i in _URLS]


_DESCRIPTION = """\
Data used to train BERTić model and its successors.
"""
_CITATION = """
@inproceedings{ljubesic-lauc-2021-bertic,
    title = "{BERT}i{\'c} - The Transformer Language Model for {B}osnian, {C}roatian, {M}ontenegrin and {S}erbian",
    author = "Ljube{\v{s}}i{\'c}, Nikola  and
      Lauc, Davor",
    editor = "Babych, Bogdan  and
      Kanishcheva, Olga  and
      Nakov, Preslav  and
      Piskorski, Jakub  and
      Pivovarova, Lidia  and
      Starko, Vasyl  and
      Steinberger, Josef  and
      Yangarber, Roman  and
      Marci{\'n}czuk, Micha{\l}  and
      Pollak, Senja  and
      P{\v{r}}ib{\'a}{\v{n}}, Pavel  and
      Robnik-{\v{S}}ikonja, Marko",
    booktitle = "Proceedings of the 8th Workshop on Balto-Slavic Natural Language Processing",
    month = apr,
    year = "2021",
    address = "Kiyv, Ukraine",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.bsnlp-1.5",
    pages = "37--42",
    abstract = "In this paper we describe a transformer model pre-trained on 8 billion tokens of crawled text from the Croatian, Bosnian, Serbian and Montenegrin web domains. We evaluate the transformer model on the tasks of part-of-speech tagging, named-entity-recognition, geo-location prediction and commonsense causal reasoning, showing improvements on all tasks over state-of-the-art models. For commonsense reasoning evaluation we introduce COPA-HR - a translation of the Choice of Plausible Alternatives (COPA) dataset into Croatian. The BERTi{\'c} model is made available for free usage and further task-specific fine-tuning through HuggingFace.",
}"""


class BerticDataConfig(datasets.BuilderConfig):
    """BuilderConfig for Bertic data sample."""

    def __init__(self, *args, subsets, **kwargs):
        """BuilderConfig for BerticData.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(BerticDataConfig, self).__init__(**kwargs)
        self.subsets = subsets


class BerticData(datasets.GeneratorBasedBuilder):
    """Bertic dataset, used for training Bertic model."""


    VERSION = datasets.Version("1.0.0")

    # This is an example of a dataset with multiple configurations.
    # If you don't want/need to define several sub-sets in your dataset,
    # just remove the BUILDER_CONFIG_CLASS and the BUILDER_CONFIGS attributes.

    # If you need to make complex sub-parts in the datasets with configurable options
    # You can create your own builder configuration class to store attribute, inheriting from BerticDataConfig
    # BUILDER_CONFIG_CLASS = MyBuilderConfig

    # You will be able to load one or the other configurations in the following list with
    # data = datasets.load_dataset('my_dataset', 'first_domain')
    # data = datasets.load_dataset('my_dataset', 'second_domain')
    BUILDER_CONFIGS = [
        BerticDataConfig(
            name="default",
            subsets=["arxiv", "open-web-math", "algebraic-stack"],
            name="default",
            subsets=["arxiv", "open-web-math", "algebraic-stack"],
            version=VERSION,
            description="All subsets",
        )
    ]
            description="All subsets",
        )
    ]

    def _info(self):
        features = datasets.Features(
            {
                "text": datasets.Value("string"),
            }
        )
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            homepage=_HOMEPAGE,
            citation=_CITATION,
        )

    def _split_generators(
        self, dl_manager: datasets.DownloadManager
    ) -> List[datasets.SplitGenerator]:
    def _split_generators(
        self, dl_manager: datasets.DownloadManager
    ) -> List[datasets.SplitGenerator]:
        urls_to_download = self._URLS
        urls_to_download = {i: url for i, url in enumerate(_URLS)}
        downloaded_files = dl_manager.download_and_extract(urls_to_download)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files[i]}
            )
            for i in urls_to_download.keys()
                name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files[i]}
            )
            for i in urls_to_download.keys()
        ]

    def _generate_examples(self, data_files):
        key = 0
        for name in data_files:
            with gzip.open(name, "rb") as f:
        key = 0
        for name in data_files:
            with gzip.open(name, "rb") as f:
                for line in f.readlines():
                    yield key, {"text": line.decode("uft-8").strip()}
                    key += 1