Medical_questions_pairs

ds = tfds.load('huggingface:medical_questions_pairs')

This dataset consists of 3048 similar and dissimilar medical question pairs hand-generated and labeled by Curai's doctors.

{
    "dr_id": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "question_1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            0,
            1
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}