SampleClient.py

import os
from openai import OpenAI
import time

# 修正 1：將 BASE_URL 設為官方地址
BASE_URL = "https://hsuwill000-llamacpp-webui01.hf.space/v1"
MODEL_ID = "LFM2-8B-A1B-UD-Q4_K_XL" # 官方 API 需要指定模型
TIMEOUT_SECONDS = 60

try:
    # 修正 2：使用同步客戶端 (OpenAI)
    client = OpenAI(
        base_url=BASE_URL,
        api_key="sk-local",  # llama.cpp 不檢查內容，只要有就行
        timeout=600
    )

    start_time = time.time()
    response = client.chat.completions.create(
        model=MODEL_ID,
        messages=[
            {"role": "system", "content": "You are a helpful and brief assistant."},
            {"role": "user", "content": "hello"}
        ]
    )

    end_time = time.time()
    
    # 輸出結果
    response_content = response.choices[0].message.content
    
    print("\n--- 🤖 模型回覆 ---")
    print(f"✅ 請求成功！耗時: {end_time - start_time:.2f} 秒")
    print(response_content.strip())
    print("--------------------")

except Exception as e:
    print(f"\n❌ 發生錯誤: {type(e).__name__}: {e}")