OpenBMB/VoxCPM

VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning

★ 38,260Forks 4,333PythonGitHub ↗Compare

Project website ↗

audio deeplearning minicpm multilingual python pytorch speech speech-synthesis text-to-speech tts tts-model voice-cloning voice-design voxcpm

Issues

Generated-audio rights clarification for VoxCPM2 2.0.3

#409 · open · 0 comments

Clarification on commercial use and long-term reuse of VoxCPM2 Voice Design outputs

#407 · open · 1 comments

Full fine-tuning on ~4K hours (1.87M samples) is taking 10 days to reach step 0 — how to speed up the full-finetuning process?

#406 · open · 3 comments

Generated audio sounds compressed with crackling/clipping artifacts on many voices

#404 · open · 0 comments

两个问题

#402 · open · 3 comments

voice tts

#401 · open · 1 comments

分享一个简洁的 TTS 桌面配音工具 yovoice，支持 macOS / Window

#400 · open · 0 comments

中文合成问题

#399 · open · 1 comments

nano-vllm-voxcpm问题反馈

#398 · open · 1 comments

Persian TTS Inference Issues: Repeated Words, Missing Words, and Inconsistent Pronunciation

#394 · open · 1 comments

[BUG] LoRA WebUI 的 seed 没有传入 `generate()`，导致推理结果无法可靠复现

#390 · open · 3 comments