llm: api_key: "YOUR_API_KEY" model: "gpt-4-turbo" #proxy: "http://<ip>:<port>" #base_url: "https://<forward_url>/v1"