msb-public
/
PaddleSpeech
mirror of https://github.com/PaddlePaddle/PaddleSpeech
1
0
Code Issues Projects Releases Wiki Activity

format

Browse Source
pull/1740/head
Hui Zhang 2 years ago
parent caf7225892
commit c7d9b11529
18 changed files with 910 additions and 823 deletions
Show Stats Download Patch File Download Diff File

2
.flake8
Unescape View File

@ -12,6 +12,8 @@ exclude =
.git, .git,
# python cache # python cache
__pycache__, __pycache__,
# third party
utils/compute-wer.py,
third_party/, third_party/,
# Provide a comma-separate list of glob patterns to include for checks. # Provide a comma-separate list of glob patterns to include for checks.
filename = filename =

4
paddlespeech/cli/asr/infer.py
Unescape View File

@ -40,6 +40,7 @@ from paddlespeech.s2t.utils.utility import UpdateConfig
__all__ = ['ASRExecutor'] __all__ = ['ASRExecutor']
@cli_register( @cli_register(
name='paddlespeech.asr', description='Speech to text infer command.') name='paddlespeech.asr', description='Speech to text infer command.')
class ASRExecutor(BaseExecutor): class ASRExecutor(BaseExecutor):
@ -278,7 +279,8 @@ class ASRExecutor(BaseExecutor):
self._outputs["result"] = result_transcripts[0] self._outputs["result"] = result_transcripts[0]
elif "conformer" in model_type or "transformer" in model_type: elif "conformer" in model_type or "transformer" in model_type:
logger.info(f"we will use the transformer like model : {model_type}") logger.info(
f"we will use the transformer like model : {model_type}")
try: try:
result_transcripts = self.model.decode( result_transcripts = self.model.decode(
audio, audio,

1
paddlespeech/server/bin/paddlespeech_client.py
Unescape View File

@ -305,6 +305,7 @@ class ASRClientExecutor(BaseExecutor):
return res['asr_results'] return res['asr_results']
@cli_client_register( @cli_client_register(
name='paddlespeech_client.cls', description='visit cls service') name='paddlespeech_client.cls', description='visit cls service')
class CLSClientExecutor(BaseExecutor): class CLSClientExecutor(BaseExecutor):

2
paddlespeech/server/engine/asr/online/ctc_search.py
Unescape View File

@ -12,7 +12,9 @@
# See the License for the specific language governing permissions and # See the License for the specific language governing permissions and
# limitations under the License. # limitations under the License.
from collections import defaultdict from collections import defaultdict
import paddle import paddle
from paddlespeech.cli.log import logger from paddlespeech.cli.log import logger
from paddlespeech.s2t.utils.utility import log_add from paddlespeech.s2t.utils.utility import log_add

2
paddlespeech/vector/cluster/diarization.py
Unescape View File

@ -20,11 +20,11 @@ A few sklearn functions are modified in this script as per requirement.
import argparse import argparse
import copy import copy
import warnings import warnings
from distutils.util import strtobool
import numpy as np import numpy as np
import scipy import scipy
import sklearn import sklearn
from distutils.util import strtobool
from scipy import linalg from scipy import linalg
from scipy import sparse from scipy import sparse
from scipy.sparse.csgraph import connected_components from scipy.sparse.csgraph import connected_components

14
speechx/examples/ngram/zh/local/text_to_lexicon.py
Unescape View File

@ -2,6 +2,7 @@
import argparse import argparse
from collections import Counter from collections import Counter
def main(args): def main(args):
counter = Counter() counter = Counter()
with open(args.text, 'r') as fin, open(args.lexicon, 'w') as fout: with open(args.text, 'r') as fin, open(args.lexicon, 'w') as fout:
@ -20,21 +21,16 @@ def main(args):
fout.write(f"{word}\t{val}\n") fout.write(f"{word}\t{val}\n")
fout.flush() fout.flush()
if __name__ == '__main__': if __name__ == '__main__':
parser = argparse.ArgumentParser( parser = argparse.ArgumentParser(
description='text(line:utt1 中国 人) to lexicon（line:中国 中 国).') description='text(line:utt1 中国 人) to lexicon（line:中国 中 国).')
parser.add_argument( parser.add_argument(
'--has_key', '--has_key', default=True, help='text path, with utt or not')
default=True,
help='text path, with utt or not')
parser.add_argument( parser.add_argument(
'--text', '--text', required=True, help='text path. line: utt1 中国 人 or 中国 人')
required=True,
help='text path. line: utt1 中国 人 or 中国 人')
parser.add_argument( parser.add_argument(
'--lexicon', '--lexicon', required=True, help='lexicon path. line:中国 中 国')
required=True,
help='lexicon path. line:中国 中 国')
args = parser.parse_args() args = parser.parse_args()
print(args) print(args)

28
speechx/examples/text_lm/local/mmseg.py
Unescape View File

@ -1,13 +1,14 @@
#!/usr/bin/env python3 #!/usr/bin/env python3
# modify from https://sites.google.com/site/homepageoffuyanwei/Home/remarksandexcellentdiscussion/page-2 # modify from https://sites.google.com/site/homepageoffuyanwei/Home/remarksandexcellentdiscussion/page-2
class Word: class Word:
def __init__(self, text='', freq=0): def __init__(self, text='', freq=0):
self.text = text self.text = text
self.freq = freq self.freq = freq
self.length = len(text) self.length = len(text)
class Chunk: class Chunk:
def __init__(self, w1, w2=None, w3=None): def __init__(self, w1, w2=None, w3=None):
self.words = [] self.words = []
@ -44,8 +45,8 @@ class Chunk:
sum += word.freq sum += word.freq
return sum return sum
class ComplexCompare:
class ComplexCompare:
def takeHightest(self, chunks, comparator): def takeHightest(self, chunks, comparator):
i = 1 i = 1
for j in range(1, len(chunks)): for j in range(1, len(chunks)):
@ -61,21 +62,25 @@ class ComplexCompare:
def mmFilter(self, chunks): def mmFilter(self, chunks):
def comparator(a, b): def comparator(a, b):
return a.totalWordLength() - b.totalWordLength() return a.totalWordLength() - b.totalWordLength()
return self.takeHightest(chunks, comparator) return self.takeHightest(chunks, comparator)
def lawlFilter(self, chunks): def lawlFilter(self, chunks):
def comparator(a, b): def comparator(a, b):
return a.averageWordLength() - b.averageWordLength() return a.averageWordLength() - b.averageWordLength()
return self.takeHightest(chunks, comparator) return self.takeHightest(chunks, comparator)
def svmlFilter(self, chunks): def svmlFilter(self, chunks):
def comparator(a, b): def comparator(a, b):
return b.standardDeviation() - a.standardDeviation() return b.standardDeviation() - a.standardDeviation()
return self.takeHightest(chunks, comparator) return self.takeHightest(chunks, comparator)
def logFreqFilter(self, chunks): def logFreqFilter(self, chunks):
def comparator(a, b): def comparator(a, b):
return a.wordFrequency() - b.wordFrequency() return a.wordFrequency() - b.wordFrequency()
return self.takeHightest(chunks, comparator) return self.takeHightest(chunks, comparator)
@ -83,6 +88,7 @@ class ComplexCompare:
dictWord = {} dictWord = {}
maxWordLength = 0 maxWordLength = 0
def loadDictChars(filepath): def loadDictChars(filepath):
global maxWordLength global maxWordLength
fsock = open(filepath) fsock = open(filepath)
@ -90,18 +96,22 @@ def loadDictChars(filepath):
freq, word = line.split() freq, word = line.split()
word = word.strip() word = word.strip()
dictWord[word] = (len(word), int(freq)) dictWord[word] = (len(word), int(freq))
maxWordLength = len(word) if maxWordLength < len(word) else maxWordLength maxWordLength = len(word) if maxWordLength < len(
word) else maxWordLength
fsock.close() fsock.close()
def loadDictWords(filepath): def loadDictWords(filepath):
global maxWordLength global maxWordLength
fsock = open(filepath) fsock = open(filepath)
for line in fsock.readlines(): for line in fsock.readlines():
word = line.strip() word = line.strip()
dictWord[word] = (len(word), 0) dictWord[word] = (len(word), 0)
maxWordLength = len(word) if maxWordLength < len(word) else maxWordLength maxWordLength = len(word) if maxWordLength < len(
word) else maxWordLength
fsock.close() fsock.close()
#判断该词word是否在字典dictWord中 #判断该词word是否在字典dictWord中
def getDictWord(word): def getDictWord(word):
result = dictWord.get(word) result = dictWord.get(word)
@ -109,14 +119,15 @@ def getDictWord(word):
return Word(word, result[1]) return Word(word, result[1])
return None return None
#开始加载字典 #开始加载字典
def run(): def run():
from os.path import join, dirname from os.path import join, dirname
loadDictChars(join(dirname(__file__), 'data', 'chars.dic')) loadDictChars(join(dirname(__file__), 'data', 'chars.dic'))
loadDictWords(join(dirname(__file__), 'data', 'words.dic')) loadDictWords(join(dirname(__file__), 'data', 'words.dic'))
class Analysis:
class Analysis:
def __init__(self, text): def __init__(self, text):
self.text = text self.text = text
self.cacheSize = 3 self.cacheSize = 3
@ -134,11 +145,10 @@ class Analysis:
if not dictWord: if not dictWord:
run() run()
def __iter__(self): def __iter__(self):
while True: while True:
token = self.getNextToken() token = self.getNextToken()
if token == None: if token is None:
raise StopIteration raise StopIteration
yield token yield token
@ -375,6 +385,8 @@ if __name__=="__main__":
cuttest(u"好人使用了它就可以解决一些问题") cuttest(u"好人使用了它就可以解决一些问题")
cuttest(u"是因为和国家") cuttest(u"是因为和国家")
cuttest(u"老年搜索还支持") cuttest(u"老年搜索还支持")
cuttest(u"干脆就把那部蒙人的闲法给废了拉倒！RT @laoshipukong : 27日，全国人大常委会第三次审议侵权责任法草案，删除了有关医疗损害责任“举证倒置”的规定。在医患纠纷中本已处于弱势地位的消费者由此将陷入万劫不复的境地。 ") cuttest(
u"干脆就把那部蒙人的闲法给废了拉倒！RT @laoshipukong : 27日，全国人大常委会第三次审议侵权责任法草案，删除了有关医疗损害责任“举证倒置”的规定。在医患纠纷中本已处于弱势地位的消费者由此将陷入万劫不复的境地。 "
)
cuttest("2022年12月30日是星期几？") cuttest("2022年12月30日是星期几？")
cuttest("二零二二年十二月三十日是星期几？") cuttest("二零二二年十二月三十日是星期几？")

2
utils/DER.py
Unescape View File

@ -26,9 +26,9 @@ import argparse
import os import os
import re import re
import subprocess import subprocess
from distutils.util import strtobool
import numpy as np import numpy as np
from distutils.util import strtobool
FILE_IDS = re.compile(r"(?<=Speaker Diarization for).+(?=\*\*\*)") FILE_IDS = re.compile(r"(?<=Speaker Diarization for).+(?=\*\*\*)")
SCORED_SPEAKER_TIME = re.compile(r"(?<=SCORED SPEAKER TIME =)[\d.]+") SCORED_SPEAKER_TIME = re.compile(r"(?<=SCORED SPEAKER TIME =)[\d.]+")

102
utils/compute-wer.py
Unescape View File

@ -1,15 +1,18 @@
#!/usr/bin/env python3 #!/usr/bin/env python3
# -*- coding: utf-8 -*- # -*- coding: utf-8 -*-
# CopyRight WeNet Apache-2.0 License # CopyRight WeNet Apache-2.0 License
import re, sys, unicodedata
import codecs import codecs
import re
import sys
import unicodedata
remove_tag = True remove_tag = True
spacelist = [' ', '\t', '\r', '\n'] spacelist = [' ', '\t', '\r', '\n']
puncts = ['!', ',', '?', puncts = [
'、', '。', '！', '，', '；', '？', '!', ',', '?', '、', '。', '！', '，', '；', '？', '：', '「', '」', '︰', '『', '』',
'：', '「', '」', '︰', '『', '』', '《', '》'] '《', '》'
]
def characterize(string): def characterize(string):
res = [] res = []
@ -43,10 +46,12 @@ def characterize(string) :
i = j i = j
return res return res
def stripoff_tags(x): def stripoff_tags(x):
if not x: return '' if not x: return ''
chars = [] chars = []
i = 0; T=len(x) i = 0
T = len(x)
while i < T: while i < T:
if x[i] == '<': if x[i] == '<':
while i < T and x[i] != '>': while i < T and x[i] != '>':
@ -78,6 +83,7 @@ def normalize(sentence, ignore_words, cs, split=None):
new_sentence.append(x) new_sentence.append(x)
return new_sentence return new_sentence
class Calculator: class Calculator:
def __init__(self): def __init__(self):
self.data = {} self.data = {}
@ -87,6 +93,7 @@ class Calculator :
self.cost['sub'] = 1 self.cost['sub'] = 1
self.cost['del'] = 1 self.cost['del'] = 1
self.cost['ins'] = 1 self.cost['ins'] = 1
def calculate(self, lab, rec): def calculate(self, lab, rec):
# Initialization # Initialization
lab.insert(0, '') lab.insert(0, '')
@ -108,10 +115,22 @@ class Calculator :
self.space[0][0]['error'] = 'non' self.space[0][0]['error'] = 'non'
for token in lab: for token in lab:
if token not in self.data and len(token) > 0: if token not in self.data and len(token) > 0:
self.data[token] = {'all' : 0, 'cor' : 0, 'sub' : 0, 'ins' : 0, 'del' : 0} self.data[token] = {
'all': 0,
'cor': 0,
'sub': 0,
'ins': 0,
'del': 0
}
for token in rec: for token in rec:
if token not in self.data and len(token) > 0: if token not in self.data and len(token) > 0:
self.data[token] = {'all' : 0, 'cor' : 0, 'sub' : 0, 'ins' : 0, 'del' : 0} self.data[token] = {
'all': 0,
'cor': 0,
'sub': 0,
'ins': 0,
'del': 0
}
# Computing edit distance # Computing edit distance
for i, lab_token in enumerate(lab): for i, lab_token in enumerate(lab):
for j, rec_token in enumerate(rec): for j, rec_token in enumerate(rec):
@ -141,7 +160,15 @@ class Calculator :
self.space[i][j]['dist'] = min_dist self.space[i][j]['dist'] = min_dist
self.space[i][j]['error'] = min_error self.space[i][j]['error'] = min_error
# Tracing back # Tracing back
result = {'lab':[], 'rec':[], 'all':0, 'cor':0, 'sub':0, 'ins':0, 'del':0} result = {
'lab': [],
'rec': [],
'all': 0,
'cor': 0,
'sub': 0,
'ins': 0,
'del': 0
}
i = len(lab) - 1 i = len(lab) - 1
j = len(rec) - 1 j = len(rec) - 1
while True: while True:
@ -184,8 +211,11 @@ class Calculator :
elif self.space[i][j]['error'] == 'non': # starting point elif self.space[i][j]['error'] == 'non': # starting point
break break
else: # shouldn't reach here else: # shouldn't reach here
print('this should not happen , i = {i} , j = {j} , error = {error}'.format(i = i, j = j, error = self.space[i][j]['error'])) print(
'this should not happen , i = {i} , j = {j} , error = {error}'.
format(i=i, j=j, error=self.space[i][j]['error']))
return result return result
def overall(self): def overall(self):
result = {'all': 0, 'cor': 0, 'sub': 0, 'ins': 0, 'del': 0} result = {'all': 0, 'cor': 0, 'sub': 0, 'ins': 0, 'del': 0}
for token in self.data: for token in self.data:
@ -195,6 +225,7 @@ class Calculator :
result['ins'] = result['ins'] + self.data[token]['ins'] result['ins'] = result['ins'] + self.data[token]['ins']
result['del'] = result['del'] + self.data[token]['del'] result['del'] = result['del'] + self.data[token]['del']
return result return result
def cluster(self, data): def cluster(self, data):
result = {'all': 0, 'cor': 0, 'sub': 0, 'ins': 0, 'del': 0} result = {'all': 0, 'cor': 0, 'sub': 0, 'ins': 0, 'del': 0}
for token in data: for token in data:
@ -205,12 +236,15 @@ class Calculator :
result['ins'] = result['ins'] + self.data[token]['ins'] result['ins'] = result['ins'] + self.data[token]['ins']
result['del'] = result['del'] + self.data[token]['del'] result['del'] = result['del'] + self.data[token]['del']
return result return result
def keys(self): def keys(self):
return list(self.data.keys()) return list(self.data.keys())
def width(string): def width(string):
return sum(1 + (unicodedata.east_asian_width(c) in "AFW") for c in string) return sum(1 + (unicodedata.east_asian_width(c) in "AFW") for c in string)
def default_cluster(word): def default_cluster(word):
unicode_names = [unicodedata.name(char) for char in word] unicode_names = [unicodedata.name(char) for char in word]
for i in reversed(range(len(unicode_names))): for i in reversed(range(len(unicode_names))):
@ -250,9 +284,15 @@ def default_cluster(word) :
return 'Other' return 'Other'
return unicode_names[0] return unicode_names[0]
def usage(): def usage():
print("compute-wer.py : compute word error rate (WER) and align recognition results and references.") print(
print(" usage : python compute-wer.py [--cs={0,1}] [--cluster=foo] [--ig=ignore_file] [--char={0,1}] [--v={0,1}] [--padding-symbol={space,underline}] test.ref test.hyp > test.wer") "compute-wer.py : compute word error rate (WER) and align recognition results and references."
)
print(
" usage : python compute-wer.py [--cs={0,1}] [--cluster=foo] [--ig=ignore_file] [--char={0,1}] [--v={0,1}] [--padding-symbol={space,underline}] test.ref test.hyp > test.wer"
)
if __name__ == '__main__': if __name__ == '__main__':
if len(sys.argv) == 1: if len(sys.argv) == 1:
@ -370,7 +410,8 @@ if __name__ == '__main__':
array = line.strip().split() array = line.strip().split()
if len(array) == 0: continue if len(array) == 0: continue
fid = array[0] fid = array[0]
rec_set[fid] = normalize(array[1:], ignore_words, case_sensitive, split) rec_set[fid] = normalize(array[1:], ignore_words, case_sensitive,
split)
# compute error rate on the interaction of reference file and hyp file # compute error rate on the interaction of reference file and hyp file
for line in open(ref_file, 'r', encoding='utf-8'): for line in open(ref_file, 'r', encoding='utf-8'):
@ -399,12 +440,14 @@ if __name__ == '__main__':
result = calculator.calculate(lab, rec) result = calculator.calculate(lab, rec)
if verbose: if verbose:
if result['all'] != 0: if result['all'] != 0:
wer = float(result['ins'] + result['sub'] + result['del']) * 100.0 / result['all'] wer = float(result['ins'] + result['sub'] + result[
'del']) * 100.0 / result['all']
else: else:
wer = 0.0 wer = 0.0
print('WER: %4.2f %%' % wer, end=' ') print('WER: %4.2f %%' % wer, end=' ')
print('N=%d C=%d S=%d D=%d I=%d' % print('N=%d C=%d S=%d D=%d I=%d' %
(result['all'], result['cor'], result['sub'], result['del'], result['ins'])) (result['all'], result['cor'], result['sub'], result['del'],
result['ins']))
space = {} space = {}
space['lab'] = [] space['lab'] = []
space['rec'] = [] space['rec'] = []
@ -446,30 +489,37 @@ if __name__ == '__main__':
rec1 = rec2 rec1 = rec2
if verbose: if verbose:
print('===========================================================================') print(
'==========================================================================='
)
print() print()
result = calculator.overall() result = calculator.overall()
if result['all'] != 0: if result['all'] != 0:
wer = float(result['ins'] + result['sub'] + result['del']) * 100.0 / result['all'] wer = float(result['ins'] + result['sub'] + result[
'del']) * 100.0 / result['all']
else: else:
wer = 0.0 wer = 0.0
print('Overall -> %4.2f %%' % wer, end=' ') print('Overall -> %4.2f %%' % wer, end=' ')
print('N=%d C=%d S=%d D=%d I=%d' % print('N=%d C=%d S=%d D=%d I=%d' %
(result['all'], result['cor'], result['sub'], result['del'], result['ins'])) (result['all'], result['cor'], result['sub'], result['del'],
result['ins']))
if not verbose: if not verbose:
print() print()
if verbose: if verbose:
for cluster_id in default_clusters: for cluster_id in default_clusters:
result = calculator.cluster([ k for k in default_clusters[cluster_id] ]) result = calculator.cluster(
[k for k in default_clusters[cluster_id]])
if result['all'] != 0: if result['all'] != 0:
wer = float(result['ins'] + result['sub'] + result['del']) * 100.0 / result['all'] wer = float(result['ins'] + result['sub'] + result[
'del']) * 100.0 / result['all']
else: else:
wer = 0.0 wer = 0.0
print('%s -> %4.2f %%' % (cluster_id, wer), end=' ') print('%s -> %4.2f %%' % (cluster_id, wer), end=' ')
print('N=%d C=%d S=%d D=%d I=%d' % print('N=%d C=%d S=%d D=%d I=%d' %
(result['all'], result['cor'], result['sub'], result['del'], result['ins'])) (result['all'], result['cor'], result['sub'], result['del'],
result['ins']))
if len(cluster_file) > 0: # compute separated WERs for word clusters if len(cluster_file) > 0: # compute separated WERs for word clusters
cluster_id = '' cluster_id = ''
cluster = [] cluster = []
@ -480,12 +530,14 @@ if __name__ == '__main__':
token.lstrip('</').rstrip('>') == cluster_id : token.lstrip('</').rstrip('>') == cluster_id :
result = calculator.cluster(cluster) result = calculator.cluster(cluster)
if result['all'] != 0: if result['all'] != 0:
wer = float(result['ins'] + result['sub'] + result['del']) * 100.0 / result['all'] wer = float(result['ins'] + result['sub'] + result[
'del']) * 100.0 / result['all']
else: else:
wer = 0.0 wer = 0.0
print('%s -> %4.2f %%' % (cluster_id, wer), end=' ') print('%s -> %4.2f %%' % (cluster_id, wer), end=' ')
print('N=%d C=%d S=%d D=%d I=%d' % print('N=%d C=%d S=%d D=%d I=%d' %
(result['all'], result['cor'], result['sub'], result['del'], result['ins'])) (result['all'], result['cor'], result['sub'],
result['del'], result['ins']))
cluster_id = '' cluster_id = ''
cluster = [] cluster = []
# begin of cluster reached, like <Keyword> # begin of cluster reached, like <Keyword>
@ -497,4 +549,6 @@ if __name__ == '__main__':
else: else:
cluster.append(token) cluster.append(token)
print() print()
print('===========================================================================') print(
'==========================================================================='
)

53
utils/format_rsl.py
Unescape View File

@ -1,11 +1,21 @@
import os # Copyright (c) 2022 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.
import argparse import argparse
import jsonlines import jsonlines
def trans_hyp(origin_hyp, def trans_hyp(origin_hyp, trans_hyp=None, trans_hyp_sclite=None):
trans_hyp = None,
trans_hyp_sclite = None):
""" """
Args: Args:
origin_hyp: The input json file which contains the model output origin_hyp: The input json file which contains the model output
@ -27,9 +37,8 @@ def trans_hyp(origin_hyp,
line = input_dict[key] + "(" + key + ".wav" + ")" + "\n" line = input_dict[key] + "(" + key + ".wav" + ")" + "\n"
f.write(line) f.write(line)
def trans_ref(origin_ref,
trans_ref = None, def trans_ref(origin_ref, trans_ref=None, trans_ref_sclite=None):
trans_ref_sclite = None):
""" """
Args: Args:
origin_hyp: The input json file which contains the model output origin_hyp: The input json file which contains the model output
@ -53,28 +62,34 @@ def trans_ref(origin_ref,
f.write(line) f.write(line)
if __name__ == "__main__": if __name__ == "__main__":
parser = argparse.ArgumentParser(prog='format hyp file for compute CER/WER', add_help=True) parser = argparse.ArgumentParser(
prog='format hyp file for compute CER/WER', add_help=True)
parser.add_argument(
'--origin_hyp', type=str, default=None, help='origin hyp file')
parser.add_argument( parser.add_argument(
'--origin_hyp', '--trans_hyp',
type=str, type=str,
default=None, default=None,
help='origin hyp file') help='hyp file for caculating CER/WER')
parser.add_argument( parser.add_argument(
'--trans_hyp', type=str, default = None, help='hyp file for caculating CER/WER') '--trans_hyp_sclite',
parser.add_argument( type=str,
'--trans_hyp_sclite', type=str, default = None, help='hyp file for caculating CER/WER by sclite') default=None,
help='hyp file for caculating CER/WER by sclite')
parser.add_argument( parser.add_argument(
'--origin_ref', '--origin_ref', type=str, default=None, help='origin ref file')
parser.add_argument(
'--trans_ref',
type=str, type=str,
default=None, default=None,
help='origin ref file') help='ref file for caculating CER/WER')
parser.add_argument(
'--trans_ref', type=str, default = None, help='ref file for caculating CER/WER')
parser.add_argument( parser.add_argument(
'--trans_ref_sclite', type=str, default = None, help='ref file for caculating CER/WER by sclite') '--trans_ref_sclite',
type=str,
default=None,
help='ref file for caculating CER/WER by sclite')
parser_args = parser.parse_args() parser_args = parser.parse_args()
if parser_args.origin_hyp is not None: if parser_args.origin_hyp is not None:

5
utils/fst/prepare_dict.py
Unescape View File

@ -82,7 +82,10 @@ def main(args):
lexicon_table.add(word) lexicon_table.add(word)
out_n += 1 out_n += 1
print(f"Filter lexicon by unit table: filter out {in_n - out_n}, {out_n}/{in_n}") print(
f"Filter lexicon by unit table: filter out {in_n - out_n}, {out_n}/{in_n}"
)
if __name__ == '__main__': if __name__ == '__main__':
parser = argparse.ArgumentParser( parser = argparse.ArgumentParser(

Write Preview
Loading…
Cancel
Save