Advanced AI Inference System with Avatars, Video, Robotics and Skills
Project description
TERMINATORI
Advanced AI Inference System — Sistema de inferencia extensible con soporte para avatares 3D, generación de vídeo en tiempo real, robótica, síntesis de voz y sistema de skills/Aureolas.
Inicio Rápido
# Clonar
git clone https://github.com/yoqer/TERMINATORI.git
cd TERMINATORI
# Configurar
cp .env.example .env
# Editar .env con tus API keys
# Iniciar (local)
chmod +x start.sh && ./start.sh
# O con Docker
./start.sh docker
Panel web: http://localhost:80
API docs: http://localhost:8000/docs
Módulos
| Módulo | Descripción | Proveedores |
|---|---|---|
| Inferencia | Motor LLM multi-backend | Ollama, OpenAI, Grok, Anthropic, llama.cpp |
| Avatar | Generación e importación 3D | Hunyuan3D (fal.ai), VRM, Live2D, GLB, FBX |
| Vídeo | Generación de vídeo | Runway Gen-4, SVD, Kling, fal.ai, MiniMax, Luma |
| Voz | TTS + STT | ElevenLabs, Grok TTS, Kokoro, OpenAI TTS, Whisper |
| Robótica | Control de robots | Isaac Lab, dimos/ROS2, Unitree G1, RoboClaw |
| Skills | Aureolas / plugins | 11 skills integradas + custom desde archivo |
| Memoria | Persistencia SQLite | Hechos, episodios, búsqueda semántica |
| API | REST + MCP + A2A | FastAPI, OpenAPI, WebSocket streaming |
Skills / Aureolas
Sistema de plugins basado en la librería Aureolas:
| Skill | Categoría | Descripción |
|---|---|---|
| AOrAlIA | AUREAS | Alineación de valores en el output |
| AureAlIA | AUREAS | Verificación de hechos + alineación |
| Mediapedia | Veridicas | Búsqueda en Wikipedia |
| IQuestLoop | AUREAS | Razonamiento iterativo multi-paso |
| ErrorInterpretar | Razonamiento | Corrección de errores de interpretación |
| Synthetic2 | Razonamiento | Datos sintéticos (Prime Intellect) |
| GenBit | Sesgos | Detección de sesgos de género |
| Safeguard | Seguridad | Filtro GPT OSS 120B-20B |
| Aardwark | Seguridad | Moderación OpenAI |
API
# Inferencia
curl -X POST http://localhost:8000/api/v1/infer \
-H "Content-Type: application/json" \
-d '{"prompt": "Hola", "model": "ollama/llama3.2"}'
# Generar avatar 3D
curl -X POST http://localhost:8000/api/v1/avatar/generate \
-d '{"prompt": "A futuristic robot", "quality": "standard"}'
# Generar vídeo
curl -X POST http://localhost:8000/api/v1/video/generate \
-d '{"prompt": "A robot walking", "provider": "runway", "duration": 5}'
# TTS
curl -X POST http://localhost:8000/api/v1/voice/synthesize \
-d '{"text": "Hola mundo", "provider": "elevenlabs"}'
# Ejecutar skill
curl -X POST http://localhost:8000/api/v1/skills/execute \
-d '{"skill": "Safeguard", "input": "texto a verificar"}'
Despliegue
Local
./start.sh local
Docker
./start.sh docker
# Ver logs: docker compose logs -f
# Detener: docker compose down
PyPI
pip install terminatori
terminatori --help
terminatori-server --port 8000
Ecosistema
TERMINATORI forma parte del ecosistema yoqer:
- Terminator1 — Motor Kokoro TTS
- Aureolas — Sistema de skills/plugins
- TerminaTodo — Gestión de almacenamiento
- TenMiNaTor — Framework de entrenamiento
Licencia
MIT © yoqer
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file terminatori-1.0.0.tar.gz.
File metadata
- Download URL: terminatori-1.0.0.tar.gz
- Upload date:
- Size: 39.6 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.8.5
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
2eb8da278f62a5f9e3564c0201e3823eae14b026c21f55dfd5d3d96f9d1727c4
|
|
| MD5 |
093fe142fe9cd77b1b9bcb2a29da0f6e
|
|
| BLAKE2b-256 |
2f8f54fb2d8329b9116e53576e8ff6870b5583df727b873cd06bc41338d185ec
|
File details
Details for the file terminatori-1.0.0-py3-none-any.whl.
File metadata
- Download URL: terminatori-1.0.0-py3-none-any.whl
- Upload date:
- Size: 41.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.8.5
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
38bf55bcba4128d314a34ef2752fd775f719846d2265b20df72bd60f182ebcb2
|
|
| MD5 |
bdce164762eaf68591941f62d85b8946
|
|
| BLAKE2b-256 |
fa242d55cd515fa9feaf13390ea7a48cc39b87fcd814c2039a2264fe9d6a1737
|