使用 LiteLLM 或 Qwen 等 LLM API 替代 OpenAI（Swarm 中应用）

使用 LiteLLM 替代 OpenAI

from swarm import Swarm, Agent from openai import OpenAIchat_model_id = 'zhipu--GLM-4-Flash' llm = OpenAI( base_url = 'http://localhost:4000/', api_key='sk-1234',
)client = Swarm(client=llm) def transfer_to_agent_b():return agent_bagent_a = Agent(name="Agent A",instructions="You are a helpful agent.",functions=[transfer_to_agent_b],model = chat_model_id,
)  agent_b = Agent(name="Agent B",instructions="Only speak in Haikus.",model = chat_model_id,
)response = client.run(agent=agent_a,messages=[{"role": "user", "content": "I want to talk to agent B."}],
)print(response.messages[-1]["content"])

---

使用 Qwen

智谱 估计和 Qwen 类似

DASHSCOPE_API_KEY = 'sk-40d1c...184f11d4'llm = OpenAI( api_key = DASHSCOPE_API_KEY, base_url="https://dashscope.aliyuncs.com/compatible-mode/v1",  # 填写 DashScope服务的base_url
)
chat_model_id = "qwen-plus"client = Swarm(client=llm) def transfer_to_agent_b():return agent_bagent_a = Agent(name="Agent A",instructions="You are a helpful agent.",functions=[transfer_to_agent_b],model = chat_model_id,
)  agent_b = Agent(name="Agent B",instructions="Only speak in Haikus.",model = chat_model_id,
)response = client.run(agent=agent_a,messages=[{"role": "user", "content": "I want to talk to agent B."}],
)print(response.messages[-1]["content"])

---

2024-10-28（一）