Voxtlm: unified decoder-only models for consolidating speech recognition/synthesis and speech/text continuation tasks

TL;DR

VoxtLM is a unified decoder-only model capable of performing speech recognition, synthesis, and text tasks, demonstrating significant improvements and open-source availability for reproducibility.

Contribution

It introduces a multitask decoder-only model integrating speech and text, advancing speech and language processing capabilities.

Findings

Improves speech synthesis intelligibility from 28.9 to 5.6

Enhances speech quality from 2.68 to 3.90

Outperforms single-task models in speech recognition and generation

Abstract

We propose a decoder-only language model, VoxtLM, that can perform four tasks: speech recognition, speech synthesis, text generation, and speech continuation. VoxtLM integrates text vocabulary with discrete speech tokens from self-supervised speech features and uses special tokens to enable multitask learning. Compared to a single-task model, VoxtLM exhibits a significant improvement in speech synthesis, with improvements in both speech intelligibility from 28.9 to 5.6 and objective quality from 2.68 to 3.90. VoxtLM also improves speech generation and speech recognition performance over the single-task counterpart. Further, VoxtLM is trained with publicly available data and training recipes and model checkpoints are open-sourced to make fully reproducible work.