MOSS-TTSD is a spoken dialogue generation model designed for expressive multi-speaker synthesis. It features long-context modeling, flexible speaker control, and multilingual support, while enabling zero-shot voice cloning from short audio references.

AnyPod is an open-source automated podcast generation tool powered by open-source TTS models such as MOSS-TTSD. It transforms any text input (TXT/PDF) into a multi-episode, high-quality podcast.

My dockerfile for moss-ttsd.

No description available

MOSS-TTSD-KAGGLE: 专为Kaggle环境优化的双语语音对话合成系统，支持中英文文本到语音对话生成，充分利用Kaggle双T4 GPU进行高质量语音合成

for colab

No description available