README.md

metadata

license: apache-2.0

azaheadhealth

Dataset INFO

features=datasets.Features(
    {
        "text": datasets.Value("string"),
        "label": datasets.ClassLabel(num_classes=2, names=["NEGATIVE", "POSITIVE"]),
    }
),
supervised_keys=None,
task_templates=[
    TextClassification(
        text_column="text", label_column="label"
    )
]

Dataset DESCRIPTION

azaheadhealth is a custom dataset for training binary text classifiers in the public health domain.

02.05.24 - The small dataset is available. This set contains a train and test split with 160 and 24 examples respectively, at roughly 10:6 Negative:Positive examples each.