跳转到内容

Google ADK:Hello World

官方源码

py
from google.adk.agents import Agent
from google.adk.tools import google_search

root_agent = Agent(
    name="search_assistant",
    model="gemini-2.0-flash-exp",  # Or your preferred Gemini model
    instruction="You are a helpful assistant. Answer user questions using Google Search when needed.",
    description="An assistant that can search the web.",
    tools=[google_search],
)

改造成qwen的源码

py
from google.adk.agents import LlmAgent
from google.adk.models.lite_llm import LiteLlm
from google.adk.tools import google_search

# 创建使用qwen模型
root_agent = LlmAgent(
    model=LiteLlm(
        model="openai/qwen-max-latest",
        api_base="https://dashscope.aliyuncs.com/compatible-mode/v1",
        api_key="sk-xxx",
    ),
    name="天蓬元帅",
    description=("a helpful assistant."),
    instruction=("You are a helpful assistant"),
    # tools=[google_search],
)

# Google search tool is not supported for model openai/qwen-max-latest
# NotImplementedError: Live connection is not supported for openai/qwen-max-latest

运行界面:

学习笔记

  1. google_search不支持qwen模型,语音和视频模式也不支持qwen模型。其他自带工具是否能适配,待测试。

  2. 与openai sdk类似,自家模型配置方便,但是别家的模型配置就稍微麻烦。也是同样需要先配置model类型,然后再构建agent。

  3. 刚出来问题不少,比如windows下的ollama使用、网页解析结果比较丑等。