pnnbao-ump

/

VieNeu-TTS-v3-Turbo

Like 60

Hi, just a gentle follow-up on this thread (opened about 9 days ago) — no rush, and thank you again for your work.

For our internal compliance review, the two points we most need confirmed for this revision (model 75ff82a72f54d55ed389e1eeb12041d3c4bac7d4, source f56ce97ffb3731aeafed623391587a1589ecb501, vieneu 3.2.3) are:

Whether the bundled preset voices and the generated audio outputs may be used in commercial / monetized content; and
The training-data provenance and speaker rights/consent for the VieNeu-TTS-10k-ENVI dataset.

A brief public note — here or in the model card, tied to this revision — would be enough for us. If it helps, we're happy to narrow the request down to just these two points. Thank you!

Hi, thanks for your patience and for the detailed follow-up.

To address your two main points for this revision (model 75ff82a72f54d55ed389e1eeb12041d3c4bac7d4, source f56ce97ffb3731aeafed623391587a1589ecb501, vieneu 3.2.3):

1. Preset voices & commercial use
   The bundled preset voices are distributed under the same Apache-2.0 license as the rest of the repository, and may be used to generate synthetic voice-over audio for commercial and monetized content.

2. Training data provenance
   We don't publicly disclose the detailed internal data collection and processing pipeline for the training dataset. What we can confirm is that the preset voices currently shipped in this revision do not, to our knowledge, infringe on third-party rights as of the release date.

We understand this may not cover every detail needed for your internal compliance review, and we appreciate you raising these questions — they're helpful for us to think through as we continue to document the project. If you need anything further clarified for your specific use case, feel free to let us know.

Thank you — that's really helpful, and the commercial-use confirmation for the preset voices and outputs is exactly what we needed.

On the training data, we completely understand not disclosing the internal pipeline. Just one narrower point for our records: can you confirm that the speakers whose voices are represented in the shipped preset voices provided consent for their voice to be used in AI training and synthetic speech generation? A simple confirmation is enough — we don't need any dataset details.

Also, if you're open to it, adding this commercial-use confirmation to the model card (tied to this revision) would help other users as well. Thanks again for your time.

Apologies for following up once more, and thank you again for your helpful clarification.

The commercial-use confirmation for the bundled preset voices and generated outputs is very helpful for us.

For our internal compliance records, there is just one narrower point we would like to confirm before relying on the shipped preset voices: would you be able to confirm whether the speakers whose voices are represented in the shipped preset voices granted appropriate rights or consent for their voices to be used in AI training and synthetic speech generation?

A simple confirmation is enough — we do not need any dataset details or internal pipeline information.

This specific rights/consent confirmation would help us determine whether we can safely use the bundled preset voices in our project.

Also, if you are open to it, adding this commercial-use and preset-voice rights/consent clarification to the model card, tied to this revision, would be helpful for other users as well.

Thank you again for your time, and apologies for the extra follow-up.

Hi,

Thanks for following up.

To confirm specifically regarding the bundled preset voices: yes, the speakers (or rightsholders) behind the shipped preset voice assets granted appropriate rights and consent for their voice data to be used in AI training and synthetic speech generation, allowing them to be safely distributed under the Apache-2.0 license for both non-commercial and commercial synthetic audio generation.

We appreciate your feedback regarding the model card documentation — we will update the model card in a future patch to explicitly reflect this clarification.

Best regards,

Thank you very much — this confirmation is exactly what we needed for our internal compliance records.

We appreciate the clear confirmation regarding both commercial synthetic audio generation and the rights/consent for the shipped preset voices.

We will keep this discussion linked to the specific revision in our records and look forward to the future model card update.

Thanks again for your time and for your work on VieNeu-TTS.

Hi @banbencochuyen ,

Following up on the model card update we mentioned — it's now live. The card has a new "Usage Rights & Licensing FAQ" section that puts the confirmations from this thread in writing, so they're associated with the repository rather than only with a discussion:

• Apache-2.0 covers all artifacts in this repository, including the bundled preset-voice assets (speaker embeddings + reference codes), not just the weights.
• The preset voices and the audio generated from them may be used in commercial and monetized content.
• The speakers/rightsholders behind the shipped preset voices granted appropriate rights and consent for AI training and synthetic speech generation.
• The training-data pipeline remains undisclosed; the confirmations above cover the preset voices and weights released here.
• Third-party components and their obligations: MOSS-Audio-Tokenizer-Nano (Apache-2.0) and sea-g2p.

These confirmations apply to the revision you pinned (model 75ff82a7…, source f56ce97f…, vieneu 3.2.3) as well as to current releases.

On the preset-voice mismatch you reported: you were right, the card was stale. The authoritative list is always list_preset_voices() for the version you have installed — the card now says so explicitly and documents the current roster (vieneu 3.3.0: 20 preset voices, default Adam). Note the roster has grown since 3.2.3, so if the exact set of voices matters for your project, keep the SDK version pinned.

One clarification we added that wasn't in this thread but is worth flagging for your review: the consent confirmation covers the bundled preset voices only. If you clone a voice from your own reference clip, obtaining the rights to that person's voice is your responsibility.

Thanks again for the detailed questions — they made the documentation better.