>>> preds = tensor([0.2, 0.3, 0.5])
    >>> target = tensor([False, True, False])
    >>> _retrieval_reciprocal_rank(preds, target)
    tensor(0.5000)

    r