grain_lg.py

import datasets

class LugandaSpeechDataset(datasets.GeneratorBasedBuilder):
    def _info(self):
        return datasets.DatasetInfo(
            description="This dataset contains speech recordings in Luganda.",
            features=datasets.Features({
                "sentence": datasets.Value("string"),
                "language": datasets.Value("string"),
                "contributor_id": datasets.Value("int64"),
                "gender": datasets.Value("string"),
                "age_group": datasets.Value("string"),
                "voice_clip": datasets.Value("string"),
                "duration": datasets.Value("float64"),
                "up_votes": datasets.Value("int64"),
                "down_votes": datasets.Value("int64"),
                "Region": datasets.Value("string"),
                "path": datasets.Audio(sampling_rate=16000),
            }),
            supervised_keys=None,
            homepage="your_dataset_homepage",
            citation="Your citation",
            languages=["lg"]
        )

    def _split_generators(self, dl_manager):
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"filepath": "data/train-*"},
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                gen_kwargs={"filepath": "data/eval-*"},
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={"filepath": "data/test-*"},
            ),
        ]

    def _generate_examples(self, filepath):
        # Your logic to parse the data file and generate examples
        with open(filepath, 'r', encoding='utf-8') as f:
            for line in f:
                data = line.strip().split(',')
                yield data[0], {
                    "sentence": data[1],
                    "language": data[2],
                    "contributor_id": int(data[3]),
                    "gender": data[4],
                    "age_group": data[5],
                    "voice_clip": data[6],
                    "duration": float(data[7]),
                    "up_votes": int(data[8]),
                    "down_votes": int(data[9]),
                    "Region": data[10],
                    "path": data[11]
                }