You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have switched to branch feature/split_arch and ran utils.sh build-xtts as this appears to be intended for development towards improved TTS. If i understand correctly this combination could be able to do multilanguage TTS.
If i request to Willow with "Respond in spanish. What time is it?" It pretty much does the right thing, expect for the actual TTS to stick with an english interpretation of the spanish language.
In the logs i see:
Coqui XTTS request with args: {'text': 'La hora actual es 19:58:09.', 'language': 'en', 'stream_chunk_size': 20, 'overlap_wav_len': 1024, 'temperature': 0.1, 'length_penalty': 1.0, 'repetition_penalty': 7.0, 'top_k': 50, 'top_p': 0.8, 'do_sample': True, 'speed': 1.0, 'enable_text_splitting': True, 'decoder': 'ne_hifigan', 'speaker': 'default'}
So it appears that by 'language': 'en', this is whats actually being requested from Coqui XTTS.
When i switch to language: str = "es" and detect_language: bool = False this stays the same.
Is this current state of in dev branch or am i simply doing somethign wrong ?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I have switched to branch feature/split_arch and ran utils.sh build-xtts as this appears to be intended for development towards improved TTS. If i understand correctly this combination could be able to do multilanguage TTS.
If i request to Willow with "Respond in spanish. What time is it?" It pretty much does the right thing, expect for the actual TTS to stick with an english interpretation of the spanish language.
In the logs i see:
Coqui XTTS request with args: {'text': 'La hora actual es 19:58:09.', 'language': 'en', 'stream_chunk_size': 20, 'overlap_wav_len': 1024, 'temperature': 0.1, 'length_penalty': 1.0, 'repetition_penalty': 7.0, 'top_k': 50, 'top_p': 0.8, 'do_sample': True, 'speed': 1.0, 'enable_text_splitting': True, 'decoder': 'ne_hifigan', 'speaker': 'default'}
So it appears that by 'language': 'en', this is whats actually being requested from Coqui XTTS.
When i switch to language: str = "es" and detect_language: bool = False this stays the same.
Is this current state of in dev branch or am i simply doing somethign wrong ?
Thanks for this impressive project.
Beta Was this translation helpful? Give feedback.
All reactions