christian.schwarz/VibeVoice
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update asr eval results

Browse Source
This commit is contained in:
YaoyaoChang
2026-01-21 09:50:24 -08:00
parent c9c778cc58
commit f7c6d2dec9
5 changed files with 7 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Figures/DER.jpg
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 61 KiB

Figures/cpWER.jpg
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 67 KiB

Figures/tcpWER.jpg
BIN
View File
Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 63 KiB

README.md
-3
View File
@@ -20,9 +20,6 @@
<h3>📰 News</h3> <h3>📰 News</h3>
<img src="https://img.shields.io/badge/Status-New-brightgreen?style=flat" alt="New" />
<img src="https://img.shields.io/badge/Feature-Realtime_TTS-blue?style=flat&logo=soundcharts" alt="Realtime TTS" />
<strong>2026-01-21: 📣 We open-sourced <a href="docs/vibevoice-asr.md"><strong>VibeVoice-ASR</strong></a>, a unified speech-to-text model designed to handle 60-minute long-form audio in a single pass, generating structured transcriptions containing Who (Speaker), When (Timestamps), and What (Content), with support for User-Customized Context.</strong> <strong>2026-01-21: 📣 We open-sourced <a href="docs/vibevoice-asr.md"><strong>VibeVoice-ASR</strong></a>, a unified speech-to-text model designed to handle 60-minute long-form audio in a single pass, generating structured transcriptions containing Who (Speaker), When (Timestamps), and What (Content), with support for User-Customized Context.</strong>
2025-12-16: 📣 We added more experimental speakers for exploration, including multilingual voices and 11 distinct English style voices. [Try it](docs/vibevoice-realtime-0.5b.md#optional-more-experimental-voices). More speaker types will be added over time. 2025-12-16: 📣 We added more experimental speakers for exploration, including multilingual voices and 11 distinct English style voices. [Try it](docs/vibevoice-realtime-0.5b.md#optional-more-experimental-voices). More speaker types will be added over time.
docs/vibevoice-asr.md
+7
View File
@@ -21,6 +21,13 @@ It is a unified speech-to-text model designed to handle **1-hour long-form audio
<img src="../Figures/VibeVoice_ASR_archi.png" alt="VibeVoice ASR Architecture" width="80%"> <img src="../Figures/VibeVoice_ASR_archi.png" alt="VibeVoice ASR Architecture" width="80%">
</p> </p>
## Evaluation
<p align="center">
<img src="../Figures/DER.jpg" alt="DER" width="80%">
<img src="../Figures/cpWER.jpg" alt="cpWER" width="80%">
<img src="../Figures/tcpWER.jpg" alt="tcpWER" width="80%">
</p>
## Installation ## Installation
We recommend to use NVIDIA Deep Learning Container to manage the CUDA environment. We recommend to use NVIDIA Deep Learning Container to manage the CUDA environment.