squad
allennlp_models.rc.tools.squad
Functions taken from the official evaluation script for SQuAD version 2.0.
make_qid_to_has_ans#
def make_qid_to_has_ans(dataset)
normalize_answer#
def normalize_answer(s)
Lower text and remove punctuation, articles and extra whitespace.
get_tokens#
def get_tokens(s)
compute_exact#
def compute_exact(a_pred: str, a_gold: str) -> int
compute_f1#
def compute_f1(a_pred: str, a_gold: str) -> float
metric_max_over_ground_truths#
def metric_max_over_ground_truths(
metric_fn: Callable[[_P, _G], _T],
prediction: _P,
ground_truths: Sequence[_G]
) -> _T
get_metric_score#
def get_metric_score(
prediction: str,
gold_answers: Sequence[str]
) -> Tuple[int, float]