Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

feat(xtts): support hindi for sentence-splitting and fine-tuning 

The XTTS model itself already supports Hindi, it was just in these components.
This commit is contained in:
Enno Hermann 2024-04-08 12:06:45 +02:00
parent e626a29251
commit d41686502e
3 changed files with 26 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
TTS/demos/xtts_ft_demo/xtts_demo.py
View File

@ -192,6 +192,7 @@ if __name__ == "__main__":
"hu", "hu",
"ko", "ko",
"ja", "ja",
"hi",
], ],
) )
progress_data = gr.Label(label="Progress:") progress_data = gr.Label(label="Progress:")
@ -370,6 +371,7 @@ if __name__ == "__main__":
"hu", "hu",
"ko", "ko",
"ja", "ja",
"hi",
], ],
) )
tts_text = gr.Textbox( tts_text = gr.Textbox(

7
TTS/tts/layers/xtts/tokenizer.py
View File

@ -11,6 +11,7 @@ from num2words import num2words
from spacy.lang.ar import Arabic from spacy.lang.ar import Arabic
from spacy.lang.en import English from spacy.lang.en import English
from spacy.lang.es import Spanish from spacy.lang.es import Spanish
from spacy.lang.hi import Hindi
from spacy.lang.ja import Japanese from spacy.lang.ja import Japanese
from spacy.lang.zh import Chinese from spacy.lang.zh import Chinese
from tokenizers import Tokenizer from tokenizers import Tokenizer
@ -19,6 +20,7 @@ from TTS.tts.layers.xtts.zh_num2words import TextNorm as zh_num2words
def get_spacy_lang(lang): def get_spacy_lang(lang):
"""Return Spacy language used for sentence splitting."""
if lang == "zh": if lang == "zh":
return Chinese() return Chinese()
elif lang == "ja": elif lang == "ja":
@ -27,8 +29,10 @@ def get_spacy_lang(lang):
return Arabic() return Arabic()
elif lang == "es": elif lang == "es":
return Spanish() return Spanish()
elif lang == "hi":
return Hindi()
else: else:
# For most languages, Enlish does the job # For most languages, English does the job
return English() return English()
@ -611,6 +615,7 @@ class VoiceBpeTokenizer:
"ja": 71, "ja": 71,
"hu": 224, "hu": 224,
"ko": 95, "ko": 95,
"hi": 150,
} }
@cached_property @cached_property

21
docs/source/models/xtts.md
View File

@ -14,16 +14,31 @@ There is no need for an excessive amount of training data that spans countless h
### Updates with v2 ### Updates with v2
- Improved voice cloning. - Improved voice cloning.
- Voices can be cloned with a single audio file or multiple audio files, without any effect on the runtime. - Voices can be cloned with a single audio file or multiple audio files, without any effect on the runtime.
- 2 new languages: Hungarian and Korean.
- Across the board quality improvements. - Across the board quality improvements.
### Code ### Code
Current implementation only supports inference and GPT encoder training. Current implementation only supports inference and GPT encoder training.
### Languages ### Languages
As of now, XTTS-v2 supports 16 languages: English (en), Spanish (es), French (fr), German (de), Italian (it), Portuguese (pt), Polish (pl), Turkish (tr), Russian (ru), Dutch (nl), Czech (cs), Arabic (ar), Chinese (zh-cn), Japanese (ja), Hungarian (hu) and Korean (ko). XTTS-v2 supports 17 languages:
Stay tuned as we continue to add support for more languages. If you have any language requests, please feel free to reach out. - Arabic (ar)
- Chinese (zh-cn)
- Czech (cs)
- Dutch (nl)
- English (en)
- French (fr)
- German (de)
- Hindi (hi)
- Hungarian (hu)
- Italian (it)
- Japanese (ja)
- Korean (ko)
- Polish (pl)
- Portuguese (pt)
- Russian (ru)
- Spanish (es)
- Turkish (tr)
### License ### License
This model is licensed under [Coqui Public Model License](https://coqui.ai/cpml). This model is licensed under [Coqui Public Model License](https://coqui.ai/cpml).