code_generation_lite.py

"""
Modified from the original code at https://huggingface.co/datasets/codeparrot/apps/blob/main/apps.py
"""

import json
import datasets


_REPO_NAME = "loubnabnl/apps"

_CITATION = """\
@article{jain2024livecodebench,
    title={LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code},
    author={Jain, Naman and Han, King and Gu, Alex and Li, Wen-Ding and Yan, Fanjia and Zhang, Tianjun and Wang, Sida and Solar-Lezama, Armando and Sen, Koushik and Stoica, Ion},
    journal={arXiv preprint arXiv:2403.07974},
    year={2024}
}
"""

_DESCRIPTION = """\
LiveCodeBench is a temporaly updating benchmark for code generation. Please check the homepage: https://livecodebench.github.io/.
"""

_HOMEPAGE = "https://livecodebench.github.io/"
_VERSIONS = ["release_v1", "release_v2"]
_VERSIONS_CONFIGS = _VERSIONS
_URLS = {
    "train": [],
    "test": ["test.jsonl", "test2.jsonl"],
}
ALLOWED_FILES = {
    "release_v1": ["test.jsonl"],
    "release_v2": ["test.jsonl", "test2.jsonl"],
}


class LCBCodeGenConfig(datasets.BuilderConfig):
    """BuilderConfig for the LCBCodeGenConfig dataset."""

    def __init__(self, *args, version_tag="release_v1", **kwargs):
        """BuilderConfig for the LCBCodeGenConfig dataset.
        Args:
            version (:obj:`List[str]`): The version of the dataset to use (only single length lists are supports).
            **kwargs: keyword arguments forwarded to super.
        """
        super().__init__(
            *args,
            name=version_tag,
            **kwargs,
        )

        assert (
            version_tag in _VERSIONS_CONFIGS
        ), f"{version_tag} not in {_VERSIONS_CONFIGS}."

        self.version_tag = version_tag


class LCBCodeGen(datasets.GeneratorBasedBuilder):
    """LCBCodeGen dataset."""

    VERSION = datasets.Version("1.0.0")

    BUILDER_CONFIG_CLASS = LCBCodeGenConfig
    BUILDER_CONFIGS = [
        LCBCodeGenConfig(version_tag=version) for version in _VERSIONS_CONFIGS
    ]
    DEFAULT_CONFIG_NAME = "release_v1"

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "question_title": datasets.Value("string"),
                    "question_content": datasets.Value("string"),
                    "platform": datasets.Value("string"),
                    "question_id": datasets.Value("string"),
                    "contest_id": datasets.Value("string"),
                    "contest_date": datasets.Value("string"),
                    "starter_code": datasets.Value("string"),
                    "difficulty": datasets.Value("string"),
                    "public_test_cases": datasets.Value("string"),
                    "private_test_cases": datasets.Value("string"),
                    "metadata": datasets.Value("string"),
                }
            ),
            supervised_keys=None,
            citation=_CITATION,
            homepage=_HOMEPAGE,
            license="MIT License",
        )

    def _split_generators(self, dl_manager):
        downloaded_files = dl_manager.download_and_extract(_URLS)

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "file_paths": downloaded_files["train"],
                    "file_names": _URLS["train"],
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "file_paths": downloaded_files["test"],
                    "file_names": _URLS["test"],
                },
            ),
        ]

    def _generate_examples(self, file_paths, file_names):
        key = 0
        for file_path, file_name in zip(file_paths, file_names):
            if file_name not in ALLOWED_FILES[self.config.version_tag]:
                continue
            for idx, line in enumerate(open(file_path, "r")):
                line_data = json.loads(line)
                yield key, line_data
                key += 1