[GH-ISSUE #2057] African Language Phonemes #1390

Open
opened 2026-10-20 10:08:42 +00:00 by bsakaguchi · 3 comments
Owner

Originally created by @paulerbear on GitHub (May 14, 2024).
Original GitHub issue: https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/2057

Hi!

I'm trying to train RVC models for a range of dialects in my country. I live in South Africa.

I'm struggling with click-consonant phonemes.

These African language sounds are not understood as phonemic by the model, so transforming voice-to-voice using recordings of African language speakers results in dropped phonemes or incorrect phonemes.

Click consonants are not reliably preserved in the output and are even occasionally replaced by other consonants ("c", "ck", "k").

Is there anyway for me to overcome this roadblock? Do I need to include more of the click-consonants in my training data?

Click Consonant Transformation.zip

Originally created by @paulerbear on GitHub (May 14, 2024). Original GitHub issue: https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/2057 Hi! I'm trying to train RVC models for a range of dialects in my country. I live in South Africa. I'm struggling with click-consonant phonemes. These African language sounds are not understood as phonemic by the model, so transforming voice-to-voice using recordings of African language speakers results in dropped phonemes or incorrect phonemes. Click consonants are not reliably preserved in the output and are even occasionally replaced by other consonants ("c", "ck", "k"). Is there anyway for me to overcome this roadblock? Do I need to include more of the click-consonants in my training data? [Click Consonant Transformation.zip](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/files/15306717/Click.Consonant.Transformation.zip)
Author
Owner

@fumiama commented on GitHub (May 26, 2024):

@RVC-Boss

<!-- gh-comment-id:2132079328 --> @fumiama commented on GitHub (May 26, 2024): @RVC-Boss
Author
Owner

@blaisewf commented on GitHub (May 26, 2024):

try training the model without pretraineds and adding more data

<!-- gh-comment-id:2132386090 --> @blaisewf commented on GitHub (May 26, 2024): try training the model without pretraineds and adding more data
Author
Owner

@arbazw commented on GitHub (Jun 21, 2024):

Add more examples of click consonants in your dataset and modify the tokenizer to treat click consonants as distinct units.

<!-- gh-comment-id:2183363606 --> @arbazw commented on GitHub (Jun 21, 2024): Add more examples of click consonants in your dataset and modify the tokenizer to treat click consonants as distinct units.
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
RVC-Project/Retrieval-based-Voice-Conversion-WebUI#1390
No description provided.