title: README
emoji: 🌖
colorFrom: green
colorTo: indigo
sdk: static
pinned: false
👽🧙 Ollama Agent Roll Cage (OARC) 🤬🤖
🌌 About OARC
Ollama Agent Roll Cage (OARC) is a Python-based framework combining the power of ollama LLMs, Coqui-TTS, Keras classifiers, LLaVA Vision, Whisper Speech Recognition, and YOLOv8 Object Detection into a unified chatbot agent API for local, custom automation. OARC enables seamless management and creation of intelligent agents through features such as:
- Conversation History Management
- Model Management and Function Calling
- Retrieval-Augmented Generation (RAG)
- Document Database Integration (via embedding models)
- Multimodal Chain of Thought
By leveraging DuckDB, external tools, and interaction spaces, OARC empowers users to structure a "base reality" for agents, enhancing control over outcomes and fostering reliable, secure automation.
OARC also offers advanced tools for creating secure agent systems. By limiting model outputs to specific training data, users can implement their own fine-tuned models while exploring logical and ontological AI frameworks.
Let’s redefine the possibilities of AI-driven agents and collaborative development! 🌟
💡 Core Features
- Unified API: Streamline chatbot agent design and deployment.
- Multimodal Capabilities: Integrate speech, vision, and data retrieval seamlessly.
- Custom Automation: Build tailored workflows for your unique use cases.
- Security Tools: Protect your agentic projects with robust safeguards.
🎥 Demo Videos
- OARC v0.28.0 - Speech-to-Speech with Vision & Agent Library
- OARC Demo Videos Compilation
- OARC 0.27.0 DEMO 5 - HF S_Dogg Model
🌍 Join the Community
- Discord: OARC | Ollama
- Support: Buy Me a Coffee
📜 Licenses
Apache License 2.0 |MADE WITH META LLAMA | Coqui Public Model License 1.0.0