TTS-1 Technical Report
This work addresses the problem of generating high-quality, expressive, and efficient speech synthesis for applications like real-time and on-device use, with support for multiple languages and emotional control.
The paper introduces Inworld TTS-1, a set of Transformer-based autoregressive text-to-speech models, including TTS-1-Max with 8.8B parameters for high quality and TTS-1 with 1.6B parameters for real-time use, achieving state-of-the-art performance on benchmarks by scaling compute and using pre-training, fine-tuning, and RL-alignment.
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.