genmix50k.py

import os
import json
import datasets

_DESCRIPTION = """
"""

_CITATION = """
"""

_LICENCE = """
"""

_FEATURES = datasets.Features({
    "x": datasets.Value(dtype="string"),
    "y": datasets.Value(dtype="string")
})

class GenMix50kConfig(datasets.BuilderConfig):
    def __init__(self, features, citation, **kwargs):
        super(GenMix50kConfig, self).__init__(version=datasets.Version("1.0.3"), **kwargs)
        self.features = features
        self.citation = citation

class GenMix50k(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        GenMix50kConfig(
            name="translation",
            data_dir="./translation",
            features=_FEATURES,
            citation=_CITATION
        ),
        GenMix50kConfig(
            name="dialogue",
            data_dir="./dialogue",
            features=_FEATURES,
            citation=_CITATION
        ),
        GenMix50kConfig(
            name="summarization",
            data_dir="./summarization",
            features=_FEATURES,
            citation=_CITATION
        )
    ]

    def _info(self) -> datasets.DatasetInfo:
        """Returns the dataset metadata."""
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=self.config.features,
            citation=self.config.citation,
            license=_LICENCE,
            supervised_keys=None
        )

    def _split_generators(self, dl_manager: datasets.DownloadManager):
        """Returns SplitGenerators"""
        data_dir = os.path.join(dl_manager.manual_dir, self.config.name)
        path_kv = {
            datasets.Split.TRAIN: os.path.join(data_dir, "train.json"),
            datasets.Split.VALIDATION: os.path.join(data_dir, "valid.json"),
            datasets.Split.TEST: os.path.join(data_dir, "test.json")
        }

        splits = []
        for split, path in path_kv.items():
            if os.path.isfile(path):
                splits.append(datasets.SplitGenerator(name=split, gen_kwargs={"filepath": path}))
            else:
                print(f"Warning: File not found for {split} split: {path}")

        if not splits:
            raise datasets.data_files.DataFilesNotFoundError(f"No valid data files found in {data_dir} for splits: {path_kv.keys()}")

        return splits

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