GLM-TTS用10万小时数据实现高效高保真语音合成，支持精准控制。
GLM-TTS Technical Report
- 两阶段架构：文本转符号+符号转波形扩散模型
- 仅用10万小时数据达开源基准最优性能
- 支持实时部署与可控语音定制，适合工业应用
本文提出GLM-TTS，一个面向生产环境的语音合成系统，兼顾效率、可控性与高保真度。该系统采用两阶段架构：先通过自回归文本到符号模型生成中间表示，再利用扩散模型将符号转换为波形。仅使用10万小时训练数据，即在多个开源基准上达到领先水平。为提升语音质量，引入带基频约束的优化语音分词器，并采用基于GRPO的多奖励强化学习框架，联合优化发音准确性、说话人相似度与表达性语调。同时，通过参数高效的LoRA语音定制和混合音素-文本输入方案，实现高效且可调控的部署。代码已开源（https://github.com/zai-org/GLM-TTS），实时合成演示可通过Z.ai (audio.z.ai) 和智谱清言应用/网站（chatglm.cn）访问。
原文摘要 · Abstract (English)
This work proposes GLM-TTS, a production-level TTS system designed for efficiency, controllability, and high-fidelity speech generation. GLM-TTS follows a two-stage architecture, consisting of a text-to-token autoregressive model and a token-to-waveform diffusion model. With only 100k hours of training data, GLM-TTS achieves state-of-the-art performance on multiple open-source benchmarks. To meet production requirements, GLM-TTS improves speech quality through an optimized speech tokenizer with fundamental frequency constraints and a GRPO-based multi-reward reinforcement learning framework that jointly optimizes pronunciation, speaker similarity, and expressive prosody. In parallel, the system enables efficient and controllable deployment via parameter-efficient LoRA-based voice customization and a hybrid phoneme-text input scheme that provides precise pronunciation control. Our code is available at https://github.com/zai-org/GLM-TTS. Real-time speech synthesis demos are provided via Z.ai (audio.z.ai), the Zhipu Qingyan app/web (chatglm.cn).
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。