Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
43 changes: 43 additions & 0 deletions .github/workflows/validate-scripts.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,43 @@
name: Validate Shell Scripts

on:
pull_request:
paths:
- 'Servicio/Instalacion-Base-VibeVoice/**/*.sh'
push:
branches:
- main
- 'feat/**'
paths:
- 'Servicio/Instalacion-Base-VibeVoice/**/*.sh'

jobs:
shellcheck:
name: ShellCheck
runs-on: ubuntu-latest
steps:
- name: Checkout code
uses: actions/checkout@v4

- name: Run ShellCheck
uses: ludeeus/action-shellcheck@master
with:
scandir: './Servicio/Instalacion-Base-VibeVoice'
severity: warning
ignore_paths: 'node_modules .git'

- name: Check script permissions
run: |
echo "Verificando permisos ejecutables en scripts..."
find Servicio/Instalacion-Base-VibeVoice -name "*.sh" -type f ! -perm -u+x -print | while read script; do
echo "ADVERTENCIA: $script no tiene permisos ejecutables"
done

# Contar scripts sin permisos ejecutables
count=$(find Servicio/Instalacion-Base-VibeVoice -name "*.sh" -type f ! -perm -u+x | wc -l)
if [ "$count" -gt 0 ]; then
echo "ERROR: $count script(s) no tienen permisos ejecutables"
exit 1
fi

echo "✓ Todos los scripts tienen permisos ejecutables correctos"
17 changes: 16 additions & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -120,4 +120,19 @@ We do not recommend using VibeVoice in commercial or real-world applications wit

## Star History

![Star History Chart](https://api.star-history.com/svg?repos=Microsoft/vibevoice&type=date&legend=top-left)
![Star History Chart](https://api.star-history.com/svg?repos=Microsoft/vibevoice&type=date&legend=top-left)

---

### ⚠️ Nota para Usuarios con NVIDIA GTX 1060 (y GPUs con arquitectura sm_61)

Esta instalación (documentada también en Servicio/Instalacion-Base-VibeVoice/LEEME.md) se puede ejecutar en **modo host** (entorno virtual Python) en vez de Docker para mejorar compatibilidad y rendimiento en GPUs antiguas como la NVIDIA GeForce GTX 1060.

Motivación: las builds precompiladas recientes de PyTorch pueden no incluir soporte para la arquitectura CUDA `sm_61`. En nuestras pruebas, la combinación estable fue usar PyTorch `2.6.0+cu118` y ejecutar la API en el `venv` del host.

Configuración recomendada:
- `VIBE_API_MODE=host` en `.env`
- `MODEL_DEVICE=cuda`
- `MODEL_PATH=/opt/vibevoice/models/VibeVoice-Realtime-0.5B/snapshots/<hash>`

Para más detalles y pasos de reversión, ver `Servicio/Instalacion-Base-VibeVoice/LEEME.md`.
Loading