Appgithub-com-vorojar-vibevoice

Download Pinokio

Sign in Create accountJoin

Sign in Create accountJoin

VibeVoice

github.com/vorojar/vibevoiceUpdated 8mo ago

Open-source AI audiobook studio. A free, private alternative to ElevenLabs. 3 voice modes, per-sentence voice & emotion control, LLM smart character analysis, mixed-voice generation. Runs 100% locally on your GPU with zero API costs.

Loading community details…

Loading project owner…

Loading launchers…

Loading check-ins…

Community

Sort byCreate post

Ask or share about VibeVoice…

Loading...

Loading references…

Loading project activity…

Pinokio

0

Projects using this repository

No references listed yet.

Check-ins (0)

Recent commits

feat: Introduce sentence editor with regeneration functionality and state management.

vorojar8 months agocb4ee2d

feat: Implement automatic -16 LUFS loudness normalization using pyloudnorm for consistent audio volume.

vorojar8 months ago02ebef2

feat: Integrate Demucs vocal extraction into audio cleaning, preceding VAD for enhanced reference audio processing.

vorojar8 months ago79495fc

feat: Introduce optional `wetext` normalization and integrate Silero VAD for automatic silence trimming in reference audio processing.

vorojar8 months agof2833e3

feat: Implement Chinese text normalization for TTS inputs and apply it within the `_inference_call` function.

vorojar8 months ago6cc84bf

feat: Introduce a new sentence editor view with mode switching, sentence splitting, and generation progress display.

vorojar8 months agofe04d18

docs: clarify why the design model cannot be batched due to varying sampled sounds.

vorojar8 months agoc19b631

feat: add GPU warmup for clone model type with dedicated dummy audio generation.

vorojar8 months ago1ab7d8b

perf: Improve TTS inference speed with batch processing, `inference_mode`, SDPA, and GPU warm-up, and add `task.md`.

vorojar8 months agob272d14

docs: significantly expand README with detailed features, ElevenLabs comparison, and roadmap.

vorojar8 months ago7b16f35