text-dataset-sample.py

import csv
import pandas as pd
import datasets

from sklearn.model_selection import train_test_split
from datasets.tasks import TextClassification

_DATASET_LABELS = ['NEGATIVE', 'POSITIVE', 'NEUTRAL']

class Custom(datasets.GeneratorBasedBuilder):
    def _info(self):
        return datasets.DatasetInfo(
            description='',
            features=datasets.Features(
                {
                    'text': datasets.Value('string'),
                    'label': datasets.features.ClassLabel(
                        names=_DATASET_LABELS
                    ),
                }
            ),
            homepage='',
            citation='',
            task_templates=[
                TextClassification(text_column='text', label_column='label')
            ],
        )

    def _split_generators(self, dl_manager):
        data_path = dl_manager.download_and_extract('data.csv')
        records = pd.read_csv(data_path)
        train_df, val_df = train_test_split(records, test_size=0.2, random_state=42)
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={'df': train_df}
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION, gen_kwargs={'df': val_df}
            ),
        ]

    def _generate_examples(self, df):
        for id_, row in df.iterrows():
            text, label = row['text'], row['label'],
            label = _DATASET_LABELS.index(label)
            yield id_, {'text': text, 'label': label}