TTS-1 Technical Report
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911081398534144 |
|---|---|
| author | Atamanenko, Oleg Chalova, Anna Coombes, Joseph Cope, Nikki Dang, Phillip Deng, Zhifeng Du, Jimmy Ermolenko, Michael Fan, Feifan Feng, Yufei Fichter, Cheryl Filimonov, Pavel Fischer, Louis Gibbs, Kylan Gusarova, Valeria Karpik, Pavel Kottner, Andreas Assad Lee, Ian Louie, Oliver Mai, Jasmine Mamontov, Mikhail Mao, Suri Morshed, Nurullah Poletaev, Igor Radu, Florin Semernia, Dmytro Shingarev, Evgenii Sivaraja, Vikram Skirko, Peter Takhautdinov, Rinat Villahermosa, Robert Wang, Jean |
| author_facet | Atamanenko, Oleg Chalova, Anna Coombes, Joseph Cope, Nikki Dang, Phillip Deng, Zhifeng Du, Jimmy Ermolenko, Michael Fan, Feifan Feng, Yufei Fichter, Cheryl Filimonov, Pavel Fischer, Louis Gibbs, Kylan Gusarova, Valeria Karpik, Pavel Kottner, Andreas Assad Lee, Ian Louie, Oliver Mai, Jasmine Mamontov, Mikhail Mao, Suri Morshed, Nurullah Poletaev, Igor Radu, Florin Semernia, Dmytro Shingarev, Evgenii Sivaraja, Vikram Skirko, Peter Takhautdinov, Rinat Villahermosa, Robert Wang, Jean |
| contents | We introduce Inworld TTS-1, a set of two Transformer-based autoregressive text-to-speech (TTS) models. Our largest model, TTS-1-Max, has 8.8B parameters and is designed for utmost quality and expressiveness in demanding applications. TTS-1 is our most efficient model, with 1.6B parameters, built for real-time speech synthesis and on-device use cases. By scaling train-time compute and applying a sequential process of pre-training, fine-tuning, and RL-alignment of the speech-language model (SpeechLM) component, both models achieve state-of-the-art performance on a variety of benchmarks, demonstrating exceptional quality relying purely on in-context learning of the speaker's voice. Inworld TTS-1 and TTS-1-Max can generate high-resolution 48 kHz speech with low latency, and support 11 languages with fine-grained emotional control and non-verbal vocalizations through audio markups. We additionally open-source our training and modeling code under an MIT license. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_21138 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TTS-1 Technical Report Atamanenko, Oleg Chalova, Anna Coombes, Joseph Cope, Nikki Dang, Phillip Deng, Zhifeng Du, Jimmy Ermolenko, Michael Fan, Feifan Feng, Yufei Fichter, Cheryl Filimonov, Pavel Fischer, Louis Gibbs, Kylan Gusarova, Valeria Karpik, Pavel Kottner, Andreas Assad Lee, Ian Louie, Oliver Mai, Jasmine Mamontov, Mikhail Mao, Suri Morshed, Nurullah Poletaev, Igor Radu, Florin Semernia, Dmytro Shingarev, Evgenii Sivaraja, Vikram Skirko, Peter Takhautdinov, Rinat Villahermosa, Robert Wang, Jean Computation and Language Artificial Intelligence Machine Learning Sound Audio and Speech Processing We introduce Inworld TTS-1, a set of two Transformer-based autoregressive text-to-speech (TTS) models. Our largest model, TTS-1-Max, has 8.8B parameters and is designed for utmost quality and expressiveness in demanding applications. TTS-1 is our most efficient model, with 1.6B parameters, built for real-time speech synthesis and on-device use cases. By scaling train-time compute and applying a sequential process of pre-training, fine-tuning, and RL-alignment of the speech-language model (SpeechLM) component, both models achieve state-of-the-art performance on a variety of benchmarks, demonstrating exceptional quality relying purely on in-context learning of the speaker's voice. Inworld TTS-1 and TTS-1-Max can generate high-resolution 48 kHz speech with low latency, and support 11 languages with fine-grained emotional control and non-verbal vocalizations through audio markups. We additionally open-source our training and modeling code under an MIT license. |
| title | TTS-1 Technical Report |
| topic | Computation and Language Artificial Intelligence Machine Learning Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2507.21138 |