Change logic for //n<Speaker>
This commit is contained in:
parent
76f9a7e714
commit
b29f8ccee9
1 changed files with 71 additions and 70 deletions
|
|
@ -45,8 +45,8 @@ if 'gpt-3.5' in MODEL:
|
|||
elif 'gpt-4' in MODEL:
|
||||
INPUTAPICOST = .01
|
||||
OUTPUTAPICOST = .03
|
||||
BATCHSIZE = 40
|
||||
FREQUENCY_PENALTY = 0
|
||||
BATCHSIZE = 50
|
||||
FREQUENCY_PENALTY = 0.1
|
||||
|
||||
#tqdm Globals
|
||||
BAR_FORMAT='{l_bar}{bar:10}{r_bar}{bar:-10b}'
|
||||
|
|
@ -460,9 +460,12 @@ def searchThings(name, pbar):
|
|||
if '<SG説明:' in name['note']:
|
||||
totalTokens[0] += translateNote(name, r'<SG説明:(.*?)>')[0]
|
||||
totalTokens[1] += translateNote(name, r'<SG説明:(.*?)>')[1]
|
||||
if '<SGカテゴリ':
|
||||
if '<SGカテゴリ:' in name['note']:
|
||||
totalTokens[0] += translateNote(name, r'<SGカテゴリ:(.*?)>')[0]
|
||||
totalTokens[1] += translateNote(name, r'<SGカテゴリ:(.*?)>')[1]
|
||||
if '<ExtendDesc:' in name['note']:
|
||||
totalTokens[0] += translateNote(name, r'<ExtendDesc:(.*?)>')[0]
|
||||
totalTokens[1] += translateNote(name, r'<ExtendDesc:(.*?)>')[1]
|
||||
|
||||
# Count totalTokens
|
||||
totalTokens[0] += nameResponse[1][0] if nameResponse != '' else 0
|
||||
|
|
@ -503,14 +506,14 @@ def searchNames(name, pbar, context):
|
|||
|
||||
# Extract Data
|
||||
responseList = []
|
||||
responseList.append(translateGPT(name['name'], newContext, True))
|
||||
responseList.append(translateGPT(name['name'], newContext, False))
|
||||
if 'Actors' in context:
|
||||
responseList.append(translateGPT(name['profile'], '', True))
|
||||
responseList.append(translateGPT(name['nickname'], 'Reply with ONLY the '+ LANGUAGE +' translation of the NPC nickname', True))
|
||||
responseList.append(translateGPT(name['profile'], '', False))
|
||||
responseList.append(translateGPT(name['nickname'], 'Reply with ONLY the '+ LANGUAGE +' translation of the NPC nickname', False))
|
||||
|
||||
if 'Armors' in context or 'Weapons' in context:
|
||||
if 'description' in name:
|
||||
responseList.append(translateGPT(name['description'], '', True))
|
||||
responseList.append(translateGPT(name['description'], '', False))
|
||||
else:
|
||||
responseList.append(['', 0])
|
||||
if 'hint' in name['note']:
|
||||
|
|
@ -621,7 +624,7 @@ def searchCodes(page, pbar, fillList, filename):
|
|||
# Using this to keep track of 401's in a row.
|
||||
currentGroup.append(jaString)
|
||||
|
||||
# Check the next code in the list
|
||||
# Join Up 401's into single string
|
||||
if len(codeList) > i+1:
|
||||
while codeList[i+1]['code'] in [401, 405, -1]:
|
||||
codeList[i]['parameters'] = []
|
||||
|
|
@ -637,33 +640,45 @@ def searchCodes(page, pbar, fillList, filename):
|
|||
if len(codeList) <= i+1:
|
||||
break
|
||||
|
||||
# Join up 401 groups for better translation.
|
||||
# Format String
|
||||
if len(currentGroup) > 0:
|
||||
finalJAString = ''.join(currentGroup).replace('？', '?')
|
||||
oldjaString = finalJAString
|
||||
|
||||
# Check for speakers in String
|
||||
# \\n<Speaker>
|
||||
matchList = re.findall(r'(.*?([\\]+[nN][wWcC]?<(.+?)>).*)', finalJAString)
|
||||
nCase = None
|
||||
if finalJAString[0] != '\\':
|
||||
regex = r'(.*?)([\\]+[nN][wWcC]?<(.*?)>.*)'
|
||||
nCase = 0
|
||||
else:
|
||||
regex = r'(.*[\\]+[nN][wWcC]?<(.*?)>)(.*)'
|
||||
nCase = 1
|
||||
matchList = re.findall(regex, finalJAString)
|
||||
if len(matchList) > 0:
|
||||
if nCase == 0:
|
||||
nametag = matchList[0][1]
|
||||
speaker = matchList[0][2]
|
||||
elif nCase == 1:
|
||||
nametag = matchList[0][0]
|
||||
speaker = matchList[0][1]
|
||||
|
||||
# Translate Speaker
|
||||
speakerID = i
|
||||
response = getSpeaker(matchList[0][1])
|
||||
speaker = response[0]
|
||||
response = getSpeaker(speaker)
|
||||
tledSpeaker = response[0]
|
||||
totalTokens[0] += response[1][0]
|
||||
totalTokens[1] += response[1][1]
|
||||
|
||||
# Set Nametag and Remove from Final String
|
||||
nametag = matchList[0][0].replace(matchList[0][1], speaker)
|
||||
finalJAString = finalJAString.replace(matchList[0][0], '')
|
||||
finalJAString = finalJAString.replace(nametag, '')
|
||||
nametag = nametag.replace(speaker, tledSpeaker)
|
||||
|
||||
# Set dialogue
|
||||
codeList[j]['parameters'][0] = matchList[0][2]
|
||||
codeList[j]['code'] = 401
|
||||
|
||||
# Remove nametag from final string
|
||||
finalJAString = finalJAString.replace(nametag, '')
|
||||
|
||||
if nCase == 0:
|
||||
codeList[i]['parameters'] = [finalJAString + nametag]
|
||||
elif nCase == 1:
|
||||
codeList[i]['parameters'] = [nametag + finalJAString]
|
||||
|
||||
### Brackets
|
||||
matchList = re.findall\
|
||||
(r'^([\\]+[cC]\[[0-9]+\]【?(.+?)】?[\\]+[cC]\[[0-9]+\])|^(【(.+)】)', finalJAString)
|
||||
|
|
@ -768,6 +783,13 @@ def searchCodes(page, pbar, fillList, filename):
|
|||
else:
|
||||
# Grab Translated String
|
||||
translatedText = fillList[0]
|
||||
|
||||
# Remove added speaker
|
||||
if speaker != '':
|
||||
matchSpeakerList = re.findall(r'(^.+?)\s?[|:]\s?', translatedText)
|
||||
if len(matchSpeakerList) > 0:
|
||||
fullSpeaker = matchSpeakerList[0]
|
||||
translatedText = re.sub(r'(^.+?)\s?[|:]\s?', '', translatedText)
|
||||
|
||||
# Textwrap
|
||||
if FIXTEXTWRAP is True:
|
||||
|
|
@ -775,20 +797,21 @@ def searchCodes(page, pbar, fillList, filename):
|
|||
if BRFLAG is True:
|
||||
translatedText = translatedText.replace('\n', '<br>')
|
||||
|
||||
# Add Beginning Text
|
||||
### Add Var Strings
|
||||
# CL Flag
|
||||
if CLFlag:
|
||||
translatedText = '\\CL' + translatedText
|
||||
CLFlag = False
|
||||
translatedText = nametag + translatedText
|
||||
nametag = ''
|
||||
translatedText = soundEffectString + translatedText
|
||||
|
||||
# Remove added speaker
|
||||
if speaker != '':
|
||||
matchSpeakerList = re.findall(r'(^.+?)\s?[|:]\s?', translatedText)
|
||||
if len(matchSpeakerList) > 0:
|
||||
fullSpeaker = matchSpeakerList[0]
|
||||
translatedText = re.sub(r'(^.+?)\s?[|:]\s?', '', translatedText)
|
||||
# Nametag
|
||||
if nCase == 0:
|
||||
translatedText = translatedText + nametag
|
||||
else:
|
||||
translatedText = nametag + translatedText
|
||||
nametag = ''
|
||||
|
||||
# //SE[#]
|
||||
translatedText = soundEffectString + translatedText
|
||||
|
||||
# Set Data
|
||||
if speakerID != None:
|
||||
|
|
@ -832,7 +855,7 @@ def searchCodes(page, pbar, fillList, filename):
|
|||
for match in matchList:
|
||||
# Remove Textwrap
|
||||
match = match.replace('\\n', ' ')
|
||||
response = translateGPT(match, 'Reply with the '+ LANGUAGE +' translation of the NPC name.', True)
|
||||
response = translateGPT(match, 'Reply with the '+ LANGUAGE +' translation of the NPC name.', False)
|
||||
translatedText = response[0]
|
||||
totalTokens[0] += response[1][0]
|
||||
totalTokens[1] += response[1][1]
|
||||
|
|
@ -1556,30 +1579,30 @@ def searchSS(state, pbar):
|
|||
if 'message1' in state:
|
||||
if len(state['message1']) > 0 and state['message1'][0] in ['は', 'を', 'の', 'に', 'が']:
|
||||
message1Response = translateGPT('Taro' + state['message1'], 'reply with only the gender neutral '+ LANGUAGE +' translation of the action log. Always start the sentence with Taro. For example,\
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', True)
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', False)
|
||||
else:
|
||||
message1Response = translateGPT(state['message1'], 'reply with only the gender neutral '+ LANGUAGE +' translation', True)
|
||||
message1Response = translateGPT(state['message1'], 'reply with only the gender neutral '+ LANGUAGE +' translation', False)
|
||||
|
||||
if 'message2' in state:
|
||||
if len(state['message2']) > 0 and state['message2'][0] in ['は', 'を', 'の', 'に', 'が']:
|
||||
message2Response = translateGPT('Taro' + state['message2'], 'reply with only the gender neutral '+ LANGUAGE +' translation of the action log. Always start the sentence with Taro. For example,\
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', True)
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', False)
|
||||
else:
|
||||
message2Response = translateGPT(state['message2'], 'reply with only the gender neutral '+ LANGUAGE +' translation', True)
|
||||
message2Response = translateGPT(state['message2'], 'reply with only the gender neutral '+ LANGUAGE +' translation', False)
|
||||
|
||||
if 'message3' in state:
|
||||
if len(state['message3']) > 0 and state['message3'][0] in ['は', 'を', 'の', 'に', 'が']:
|
||||
message3Response = translateGPT('Taro' + state['message3'], 'reply with only the gender neutral '+ LANGUAGE +' translation of the action log. Always start the sentence with Taro. For example,\
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', True)
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', False)
|
||||
else:
|
||||
message3Response = translateGPT(state['message3'], 'reply with only the gender neutral '+ LANGUAGE +' translation', True)
|
||||
message3Response = translateGPT(state['message3'], 'reply with only the gender neutral '+ LANGUAGE +' translation', False)
|
||||
|
||||
if 'message4' in state:
|
||||
if len(state['message4']) > 0 and state['message4'][0] in ['は', 'を', 'の', 'に', 'が']:
|
||||
message4Response = translateGPT('Taro' + state['message4'], 'reply with only the gender neutral '+ LANGUAGE +' translation of the action log. Always start the sentence with Taro. For example,\
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', True)
|
||||
Translate \'Taroを倒した！\' as \'Taro was defeated!\'', False)
|
||||
else:
|
||||
message4Response = translateGPT(state['message4'], 'reply with only the gender neutral '+ LANGUAGE +' translation', True)
|
||||
message4Response = translateGPT(state['message4'], 'reply with only the gender neutral '+ LANGUAGE +' translation', False)
|
||||
|
||||
# if 'note' in state:
|
||||
if 'help' in state['note']:
|
||||
|
|
@ -1708,20 +1731,10 @@ def searchSystem(data, pbar):
|
|||
# Save some money and enter the character before translation
|
||||
def getSpeaker(speaker):
|
||||
match speaker:
|
||||
case 'アイル':
|
||||
return 'Aeru'
|
||||
case 'リラ':
|
||||
return 'Lira'
|
||||
case 'アザミ':
|
||||
return 'Azami'
|
||||
case 'マーガレット':
|
||||
return 'Margaret'
|
||||
case 'ミール':
|
||||
return 'Miiru'
|
||||
case 'ライト':
|
||||
return 'Light'
|
||||
case '雪音':
|
||||
return ['Yukine', [0,0]]
|
||||
case _:
|
||||
return [speaker, [0,0]]
|
||||
return translateGPT(speaker, 'Reply with only the '+ LANGUAGE +' translation of the NPC name.', False)
|
||||
|
||||
def subVars(jaString):
|
||||
jaString = jaString.replace('\u3000', ' ')
|
||||
|
|
@ -1844,23 +1857,11 @@ def batchList(input_list, batch_size):
|
|||
|
||||
def createContext(fullPromptFlag, subbedT):
|
||||
characters = 'Game Characters:\
|
||||
護 == Name: Mamoru - Male\
|
||||
神代 一騎 == Last Name: Kamishiro, First Name: Ikki - Male\
|
||||
神代 琴音 == Last Name: Kamishiro, First Name: Kotone - Female\
|
||||
神代 莉々子 == Last Name: Kamishiro, First Name: Ririko - Female\
|
||||
神代 紗夜 == Last Name: Kamishiro, First Name: Saya - Female\
|
||||
篠原漣 == Last Name: Shinohara, First Name: Ren - Male\
|
||||
藪井 == Name: Yabui - Male\
|
||||
舟木 == Name: Funaki - Male\
|
||||
貞二 == Name: Jouji - Male\
|
||||
兼田 響子 == Last Name: Kaneda, First Name: Kyouko - Female\
|
||||
兼田 真人 == Last Name: Kaneda, First Name: Masato - Male\
|
||||
小出 == Name: Koide - Male\
|
||||
進士 == Name: Shinji - Male\
|
||||
雪乃 == Name: Yukino - Female'
|
||||
雪音 == Yukine - Female\
|
||||
少女 == Name: Girl, Gender: Female'
|
||||
|
||||
system = PROMPT if fullPromptFlag else \
|
||||
f'Output ONLY the {LANGUAGE} translation in the following format: `Translation: <{LANGUAGE.upper()}_TRANSLATION>`'
|
||||
f'Output ONLY the {LANGUAGE} translation in the following format: `Translation: <{LANGUAGE.upper()}_TRANSLATION>`\nNever include any notes, explanations, dislaimers, or anything similar in your response\n- "Game Characters" - The names, nicknames, and genders of the game characters. Reference this to know the names, nicknames, and gender of characters in the game.\n'
|
||||
user = f'{subbedT}'
|
||||
return characters, system, user
|
||||
|
||||
|
|
@ -1930,7 +1931,7 @@ def countTokens(characters, system, user, history):
|
|||
inputTotalTokens += len(enc.encode(user))
|
||||
|
||||
# Output
|
||||
outputTotalTokens += round(len(enc.encode(user))/1.7)
|
||||
outputTotalTokens += round(len(enc.encode(user))/2)
|
||||
|
||||
return [inputTotalTokens, outputTotalTokens]
|
||||
|
||||
|
|
@ -1983,7 +1984,7 @@ def translateGPT(text, history, fullPromptFlag):
|
|||
extractedTranslations = extractTranslation(translatedTextList, True)
|
||||
tList[index] = extractedTranslations
|
||||
if len(tList[index]) != len(translatedTextList):
|
||||
print('Test')
|
||||
mismatch = True # Just here so breakpoint can be set
|
||||
history = extractedTranslations[-10:] # Update history if we have a list
|
||||
else:
|
||||
# Ensure we're passing a single string to extractTranslation
|
||||
|
|
|
|||
Loading…
Reference in a new issue