Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix zh bug #3238

Merged
merged 1 commit into from
Nov 16, 2023
Merged

Fix zh bug #3238

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 8 additions & 6 deletions TTS/tts/layers/xtts/tokenizer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -170,7 +170,7 @@ def split_sentence(text, lang, text_split_length=250):
# There are not many common abbreviations in Arabic as in English.
]
],
"zh-cn": [
"zh": [
(re.compile("\\b%s\\." % x[0], re.IGNORECASE), x[1])
for x in [
# Chinese doesn't typically use abbreviations in the same way as Latin-based scripts.
Expand Down Expand Up @@ -335,7 +335,7 @@ def expand_abbreviations_multilingual(text, lang="en"):
("°", " درجة "),
]
],
"zh-cn": [
"zh": [
# Chinese
(re.compile(r"%s" % re.escape(x[0]), re.IGNORECASE), x[1])
for x in [
Expand Down Expand Up @@ -619,14 +619,15 @@ def katsu(self):
return cutlet.Cutlet()

def check_input_length(self, txt, lang):
lang = lang.split("-")[0] # remove the region
limit = self.char_limits.get(lang, 250)
if len(txt) > limit:
print(
f"[!] Warning: The text length exceeds the character limit of {limit} for language '{lang}', this might cause truncated audio."
)

def preprocess_text(self, txt, lang):
if lang in {"ar", "cs", "de", "en", "es", "fr", "hu", "it", "nl", "pl", "pt", "ru", "tr", "zh-cn", "ko"}:
if lang in {"ar", "cs", "de", "en", "es", "fr", "hu", "it", "nl", "pl", "pt", "ru", "tr", "zh", "ko"}:
txt = multilingual_cleaners(txt, lang)
if lang == "zh":
txt = chinese_transliterate(txt)
Expand All @@ -642,6 +643,7 @@ def encode(self, txt, lang):
lang = lang.split("-")[0] # remove the region
self.check_input_length(txt, lang)
txt = self.preprocess_text(txt, lang)
lang = "zh-cn" if lang == "zh" else lang
txt = f"[{lang}]{txt}"
txt = txt.replace(" ", "[SPACE]")
return self.tokenizer.encode(txt).ids
Expand Down Expand Up @@ -738,8 +740,8 @@ def test_expand_numbers_multilingual():
("Dat wordt dan $20 meneer.", "Dat wordt dan twintig dollar meneer.", "nl"),
("Dat wordt dan 20€ meneer.", "Dat wordt dan twintig euro meneer.", "nl"),
# Chinese (Simplified)
("在12.5秒内", "在十二点五秒内", "zh-cn"),
("有50名士兵", "有五十名士兵", "zh-cn"),
("在12.5秒内", "在十二点五秒内", "zh"),
("有50名士兵", "有五十名士兵", "zh"),
# ("那将是$20先生", '那将是二十美元先生', 'zh'), currency doesn't work
# ("那将是20€先生", '那将是二十欧元先生', 'zh'),
# Turkish
Expand Down Expand Up @@ -820,7 +822,7 @@ def test_symbols_multilingual():
("Ik heb 14% batterij", "Ik heb 14 procent batterij", "nl"),
("Ik zie je @ het feest", "Ik zie je bij het feest", "nl"),
("لدي 14% في البطارية", "لدي 14 في المئة في البطارية", "ar"),
("我的电量为 14%", "我的电量为 14 百分之", "zh-cn"),
("我的电量为 14%", "我的电量为 14 百分之", "zh"),
("Pilim %14 dolu.", "Pilim yüzde 14 dolu.", "tr"),
("Az akkumulátorom töltöttsége 14%", "Az akkumulátorom töltöttsége 14 százalék", "hu"),
("배터리 잔량이 14%입니다.", "배터리 잔량이 14 퍼센트입니다.", "ko"),
Expand Down
2 changes: 1 addition & 1 deletion TTS/tts/models/xtts.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -396,7 +396,7 @@ def inference_with_config(self, text, config, ref_audio_path, language, **kwargs
inference with config
"""
assert (
language in self.config.languages
"zh-cn" if language == "zh" else language in self.config.languages
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It will return "zh-cn" when language == "zh" and it will check if not. It is ok but for both conditional returns the same data type it should be True if language == "zh" else language in self.config.languages.

), f" ❗ Language {language} is not supported. Supported languages are {self.config.languages}"
# Use generally found best tuning knobs for generation.
settings = {
Expand Down
Loading