import json import urllib.request BASE = "http://127.0.0.1:30000" def post(path, body): request = urllib.request.Request(BASE + path, json.dumps(body).encode(), {"Content-Type": "application/json"}) with urllib.request.urlopen(request, timeout=120) as response: return json.loads(response.read()) native = post("/generate", {"text": "The capital of France is", "sampling_params": {"max_new_tokens": 12, "temperature": 0}}) print("POST /generate") print(json.dumps(native, indent=2)) chat = post("/v1/chat/completions", { "model": "default", "messages": [{"role": "user", "content": "Name one planet. Answer in one word."}], "max_tokens": 64, "temperature": 0, "chat_template_kwargs": {"enable_thinking": False}, }) print("POST /v1/chat/completions") print(json.dumps(chat, indent=2))