clustering_files.py

import datasets
import pandas as pd
import numpy as np

logger = datasets.logging.get_logger(__name__)

_DATA_PATH = "https://huggingface.co/datasets/conversy/clustering_files/resolve/main/dataset.parquet"

class ClusteringFilesConfig(datasets.BuilderConfig):
    """BuilderConfig for Conversy Benchmark."""

    def __init__(self, name, version, **kwargs):
        """BuilderConfig for Conversy Benchmark.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        self.name = name
        self.version = version
        self.features = kwargs.pop("features", None)
        self.description = kwargs.pop("description", None)
        self.data_url = kwargs.pop("data_url", None)
        self.nb_data_shards = kwargs.pop("nb_data_shards", None)

        super(ClusteringFilesConfig, self).__init__(
            name=name,
            version=version,
            **kwargs
        )


class ClusteringFiles(datasets.GeneratorBasedBuilder):
    """Conversy benchmark"""
    VERSION = datasets.Version("1.0.0")
    BUILDER_CONFIGS = [
        ClusteringFilesConfig(
                name="ClusteringFiles",
                version=VERSION,
                description="Conversy Benchmark for ML models evaluation",
                features=["filename", "segments"],
                data_url=_DATA_PATH,
                nb_data_shards=1)
    ]

    def _info(self):
        description = (
            "Voice Print Clustering Benchmark"
        )
        features = datasets.Features(
            {
                "filename": datasets.Value("string"),
                "segments": [
                    {
                        "segment_id": datasets.Value("string"),
                        "speaker": datasets.Value("string"),
                        "duration": datasets.Value("float32"),
                        "segment_clean": datasets.Value("bool"),
                        "start": datasets.Value("float32"),
                        "end": datasets.Value("float32"),
                        "readable_start": datasets.Value("string"),
                        "readable_end": datasets.Value("string"),
                        "vp": datasets.Sequence(datasets.Value("float32")),
                    }
                ]
            })
        return datasets.DatasetInfo(
            description=description,
            features=features,
            supervised_keys=None,
            version=self.config.version
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        data_url = self.config.data_url
        downloaded_file = dl_manager.download_and_extract(data_url)
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"file_path": downloaded_file},
            ),
        ]

    def _generate_examples(self, file_path):
        """Yields examples."""
        df = pd.read_parquet(file_path)

        files = {}
        for idx, row in df.iterrows():
            if row["filename"] not in files:
                files[row["filename"]] = {
                    "filename": row["filename"],
                    "segments": []
                }
            files[row["filename"]]["segments"].append({
                "segment_id": row["segment_id"],
                "speaker": row.get("speaker", 'unknown'),
                "duration": row["duration"],
                "segment_clean": row.get("segment_clean", True),
                "start": row['start'],
                "end": row['end'],
                "readable_start": row['readable_start'],
                "readable_end": row['readable_end'],
                "vp": row["vp"]
            })

        for idx, file_data in enumerate(files.values()):
            for segment in file_data["segments"]:
                segment["vp"] = segment["vp"].tolist()
            yield idx, file_data