utils_qa.py

# coding=utf-8
# Copyright 2020 The HuggingFace Team All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
Post-processing utilities for question answering.
"""
import string
import re
import json
from collections import Counter
from statistics import mean

from utils import ensure_file_path


def normalize_answer(s):
    """Lower text and remove punctuation, articles and extra whitespace."""

    def remove_articles(text):
        return re.sub(r'\b(a|an|the)\b', ' ', text)

    def white_space_fix(text):
        return ' '.join(text.split())

    def remove_punc(text):
        exclude = set(string.punctuation)
        return ''.join(ch for ch in text if ch not in exclude)

    def lower(text):
        return text.lower()

    return white_space_fix(remove_articles(remove_punc(lower(s))))


def f1_score(prediction, ground_truth):
    prediction_tokens = normalize_answer(prediction).split()
    ground_truth_tokens = normalize_answer(ground_truth).split()
    common = Counter(prediction_tokens) & Counter(ground_truth_tokens)
    num_same = sum(common.values())
    if num_same == 0:
        return 0
    precision = 1.0 * num_same / len(prediction_tokens)
    recall = 1.0 * num_same / len(ground_truth_tokens)
    f1 = (2 * precision * recall) / (precision + recall)
    return f1


def exact_match_score(prediction, ground_truth):
    return normalize_answer(prediction) == normalize_answer(ground_truth)


def metric_max_over_ground_truths(metric_fn, prediction, ground_truths):
    scores_for_ground_truths = []
    for ground_truth in ground_truths:
        score = metric_fn(prediction, ground_truth)
        scores_for_ground_truths.append(score)
    return max(scores_for_ground_truths)


def calc_score(gold_answers, predictions):
    em_lst = []
    f1_lst = []

    for ground_truths, prediction in zip(gold_answers, predictions):
        em_lst.append(metric_max_over_ground_truths(exact_match_score, prediction, ground_truths))
        f1_lst.append(metric_max_over_ground_truths(f1_score, prediction, ground_truths))

    aggregated_em = mean(em_lst)
    aggregated_f1 = mean(f1_lst)

    return aggregated_em, aggregated_f1, em_lst, f1_lst


def output_pred(predictions, references, em_lst, f1_lst, output_pred_path, header_dict=None):
    assert len(predictions) == len(references) == len(em_lst) == len(f1_lst)
    ensure_file_path(output_pred_path)
    with open(output_pred_path, 'w', encoding='utf-8') as f:
        if header_dict:
            f.write(json.dumps(header_dict) + '\n')
        for pred, ref_lst, em, f1 in zip(predictions, references, em_lst, f1_lst):
            output_dict = {
                'pred': pred,
                'ref_lst': ref_lst,
                'em': em,
                'f1': f1
            }
            f.write(json.dumps(output_dict) + '\n')