BASE TTS: Lessons from building a billion-parameter Text-to-Speech model on 100K hours of data

TL;DR

BASE TTS is a large-scale, 1-billion-parameter text-to-speech model trained on 100K hours of data, achieving state-of-the-art naturalness and demonstrating emergent abilities like natural prosody on complex sentences.

Contribution

We introduce BASE TTS, the largest TTS model to date with novel speech tokenization and demonstrate emergent abilities in large-scale TTS models.

Findings

Achieved state-of-the-art speech naturalness.

Large models show emergent abilities like natural prosody.

Developed a new dataset to measure emergent TTS abilities.

Abstract

We introduce a text-to-speech (TTS) model called BASE TTS, which stands for $\textbf{B}$ig $\textbf{A}$daptive $\textbf{S}$treamable TTS with $\textbf{E}$mergent abilities. BASE TTS is the largest TTS model to-date, trained on 100K hours of public domain speech data, achieving a new state-of-the-art in speech naturalness. It deploys a 1-billion-parameter autoregressive Transformer that converts raw texts into discrete codes ("speechcodes") followed by a convolution-based decoder which converts these speechcodes into waveforms in an incremental, streamable manner. Further, our speechcodes are built using a novel speech tokenization technique that features speaker ID disentanglement and compression with byte-pair encoding. Echoing the widely-reported "emergent abilities" of large language models when trained on increasing volume of data, we show that BASE TTS variants built with 10K+ hours and 500M+ parameters begin to demonstrate natural prosody on textually complex sentences. We design and share a specialized dataset to measure these emergent abilities for text-to-speech. We showcase state-of-the-art naturalness of BASE TTS by evaluating against baselines that include publicly available large-scale text-to-speech systems: YourTTS, Bark and TortoiseTTS. Audio samples generated by the model can be heard at https://amazon-ltts-paper.com/.