You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
79 lines
2.7 KiB
79 lines
2.7 KiB
4 years ago
|
# Copyright (c) 2021 PaddlePaddle Authors. 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.
|
||
8 years ago
|
"""External Scorer for Beam Search Decoder."""
|
||
|
import os
|
||
4 years ago
|
|
||
8 years ago
|
import kenlm
|
||
|
import numpy as np
|
||
|
|
||
|
|
||
7 years ago
|
class Scorer(object):
|
||
8 years ago
|
"""External scorer to evaluate a prefix or whole sentence in
|
||
|
beam search decoding, including the score from n-gram language
|
||
|
model and word count.
|
||
8 years ago
|
|
||
8 years ago
|
:param alpha: Parameter associated with language model. Don't use
|
||
|
language model when alpha = 0.
|
||
8 years ago
|
:type alpha: float
|
||
8 years ago
|
:param beta: Parameter associated with word count. Don't use word
|
||
|
count when beta = 0.
|
||
8 years ago
|
:type beta: float
|
||
|
:model_path: Path to load language model.
|
||
4 years ago
|
:type model_path: str
|
||
8 years ago
|
"""
|
||
|
|
||
|
def __init__(self, alpha, beta, model_path):
|
||
|
self._alpha = alpha
|
||
|
self._beta = beta
|
||
|
if not os.path.isfile(model_path):
|
||
|
raise IOError("Invaid language model path: %s" % model_path)
|
||
|
self._language_model = kenlm.LanguageModel(model_path)
|
||
|
|
||
|
# n-gram language model scoring
|
||
8 years ago
|
def _language_model_score(self, sentence):
|
||
8 years ago
|
#log10 prob of last word
|
||
|
log_cond_prob = list(
|
||
|
self._language_model.full_scores(sentence, eos=False))[-1][0]
|
||
|
return np.power(10, log_cond_prob)
|
||
|
|
||
|
# word insertion term
|
||
8 years ago
|
def _word_count(self, sentence):
|
||
8 years ago
|
words = sentence.strip().split(' ')
|
||
|
return len(words)
|
||
|
|
||
8 years ago
|
# reset alpha and beta
|
||
|
def reset_params(self, alpha, beta):
|
||
|
self._alpha = alpha
|
||
|
self._beta = beta
|
||
|
|
||
8 years ago
|
# execute evaluation
|
||
|
def __call__(self, sentence, log=False):
|
||
|
"""Evaluation function, gathering all the different scores
|
||
|
and return the final one.
|
||
|
|
||
|
:param sentence: The input sentence for evalutation
|
||
4 years ago
|
:type sentence: str
|
||
8 years ago
|
:param log: Whether return the score in log representation.
|
||
|
:type log: bool
|
||
|
:return: Evaluation score, in the decimal or log.
|
||
|
:rtype: float
|
||
|
"""
|
||
8 years ago
|
lm = self._language_model_score(sentence)
|
||
|
word_cnt = self._word_count(sentence)
|
||
4 years ago
|
if log is False:
|
||
8 years ago
|
score = np.power(lm, self._alpha) * np.power(word_cnt, self._beta)
|
||
8 years ago
|
else:
|
||
8 years ago
|
score = self._alpha * np.log(lm) + self._beta * np.log(word_cnt)
|
||
8 years ago
|
return score
|