kriyans.py

import datasets

_URL = "https://colab.research.google.com/drive/1aSWUrQFWU_RJOt9NmVDRnn9ci-11dC9x#scrollTo=LOncaJUnrl8O"

# Define the path to your CSV file
csv_file_path = '/indian-name-org.csv'

class YourCustomConfig(datasets.BuilderConfig):
    def __init__(self, **kwargs):
        super(YourCustomConfig, self).__init__(**kwargs)

class YourCustomDataset(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        YourCustomConfig(
            name="your_custom_dataset",
            version=datasets.Version("1.0.0"),
            description="Your Custom Dataset",
        ),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description="Your custom dataset description",
            features=datasets.Features(
                {
                    "id": datasets.Value("string"),
                    "tokens": datasets.Sequence(datasets.Value("string")),
                    "ner_tags": datasets.Sequence(
                        datasets.features.ClassLabel(
                            names=[
                                "B-PER",
                                "I-ORG",
                            ]
                        )
                    ),
                }
            ),
            supervised_keys=None,
        )

    def _split_generators(self, dl_manager):
         urls_to_download = {
            "train": f"{_URL}{csv_file_path}",
        }
        downloaded_files = dl_manager.download_and_extract(urls_to_download)

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

    def _generate_examples(self, filepath):
        with open(filepath, encoding="utf-8") as f:
            current_tokens = []
            current_labels = []
            sentence_counter = 0
            for row in f:
                row = row.rstrip()
                if row:
                    token, label = row.split(",")
                    current_tokens.append(token)
                    current_labels.append(label)
                else:
                    if not current_tokens:
                        continue
                    assert len(current_tokens) == len(current_labels), "Mismatch between tokens and labels"
                    sentence = (
                        sentence_counter,
                        {
                            "id": str(sentence_counter),
                            "tokens": current_tokens,
                            "ner_tags": current_labels,
                        },
                    )
                    sentence_counter += 1
                    current_tokens = []
                    current_labels = []
                    yield sentence
            if current_tokens:
                yield sentence_counter, {
                    "id": str(sentence_counter),
                    "tokens": current_tokens,
                    "ner_tags": current_labels,
                }