JARVIS: A Collaborative System for Solving Complicated AI Tasks

JARVIS is a collaborative system that uses an LLM as the controller and numerous expert models as collaborative executors (from HuggingFace Hub) to solve complicated AI tasks. The workflow of the system consists of four stages: Task Planning, Model Selection, Task Execution, and Response Generation, which involve various AI models deployed locally or on Hugging Face Inference Endpoints. The system requires default and minimum Ubuntu 16.04 LTS configurations and offers various modes of inference endpoints, including local, HuggingFace, and hybrid. Users can access JARVIS' services by using the Web API, Gradio demo, CLI mode, or user-friendly web page.