gpu_occupy.py

from openai import OpenAI
import time

# ================= 配置区域 =================
API_KEY = "token-abc123"
BASE_URL = "http://127.0.0.1:1054/v1"
MODEL_NAME = "deepseek-v3.2"

client = OpenAI(api_key=API_KEY, base_url=BASE_URL)

while True:
    response = client.chat.completions.create(
        model=MODEL_NAME,
        messages=[
            {"role": "user", "content": "你好"}
        ],
        temperature=0.7,
    )
    # print(response)
    time.sleep(10)