dataset_script.py

import json
import datasets

_DESCRIPTION = """Multi-domain language sequence datasets."""

class MultiDomainDataset(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="domain_1", version=datasets.Version("1.0.0"), description="Domain 1 dataset"),
        datasets.BuilderConfig(name="domain_2", version=datasets.Version("1.0.0"), description="Domain 2 dataset"),
        datasets.BuilderConfig(name="domain_3", version=datasets.Version("1.0.0"), description="Domain 3 dataset"),
        datasets.BuilderConfig(name="domain_4", version=datasets.Version("1.0.0"), description="Domain 4 dataset"),
        datasets.BuilderConfig(name="domain_5", version=datasets.Version("1.0.0"), description="Domain 5 dataset"),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features({
                "text": datasets.Value("string"),
            }),
            supervised_keys=None,
        )

    def _split_generators(self, dl_manager):
        file_map = {
            "domain_1": "domain_0.jsonl",
            "domain_2": "domain_1.jsonl",
            "domain_3": "domain_2.jsonl",
            "domain_4": "domain_3.jsonl",
            "domain_5": "domain_4.jsonl",
        }
        data_file = dl_manager.download_and_extract(file_map[self.config.name])

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"filepath": data_file},
            )
        ]

    def _generate_examples(self, filepath):
        with open(filepath, encoding="utf-8") as f:
            for i, line in enumerate(f):
                yield i, json.loads(line)