- Published on
【论文分享】| cosyvoice语音合成论文分享
CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens
Wonderful stories from PaddlePaddle contributors
CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens
“我之前以为 GLCC 就是大厂开放一些边角料课题给在校生练练手,但参与之后发现,飞桨的赛题足够硬核,它的难度、复杂度、完备度都远超我的预期。最终,它给我的收获也远超预期。”
从文档深耕到模型适配,2 个月集训见证 29 位新 contributor 蜕变