ollama run qwen3:4b-instruct #会下载qwen3:4b-instruct大模型并运行，然后就可以问问题了
#这是我之后下载的几个，目前qwen3:4b 和 qwen3:4b-instruct 最符合电脑性能
ollama list | sort

deepseek-r1:1.5b     e0979632db5a    1.1 GB    7 hours ago
deepseek-r1:7b       755ced02ce7b    4.7 GB    6 hours ago
gemma3:4b            a2af6cc3eb7f    3.3 GB    2 hours ago
llama3.2:3b          a80c4f17acd5    2.0 GB    6 hours ago
NAME                 ID              SIZE      MODIFIED
qwen3:4b             359d7dd4bcda    2.5 GB    8 hours ago
qwen3:4b-instruct    0edcdef34593    2.5 GB    7 hours ago
qwen3:8b             500a1f067a9f    5.2 GB    5 hours ago

ollama run qwen3:4b-instruct  --hidethinking #使用hidethingking隐藏思考过程
#或者在运行中，使用/set nothink 隐藏思考过程，不过不是每个大模型都有效

sudo apt install python3-requests

nano test_llm.py

import requests

url = "http://192.168.6.201:11434/api/chat"


data = {
    "model": "qwen3:4b-instruct",
    "messages": [
        {
            "role": "user",
            "content": "你好，随机给我一个五位数"
        }
    ],
    "stream": False
}


print("发送请求")

r = requests.post(
    url,
    json=data,
    timeout=60
)

print("状态码:", r.status_code)

print(r.text)

╭─ ~                                                                ly@ubt24 18:10:49
╰─❯ python3 test_llm.py
发送请求
状态码: 200
{"model":"qwen3:4b-instruct","created_at":"2026-08-26T10:11:13.7346513Z","message":{"role":"assistant","content":"你好！这是一个随机生成的五位数：**73924** 😊\n\n如果需要更多，比如带条件的（比如是偶数、能被3整除等），也可以告诉我哦！"},"done":true,"done_reason":"stop","total_duration":7589013100,"load_duration":7398700,"prompt_eval_count":16,"prompt_eval_duration":143164000,"eval_count":46,"eval_duration":7413854000}