[Bug] Doubao LLM for running the agent team demo

[JokerHB]

Description

When replacing the LLM models with Doubao, an error occurs.
However, it runs well when using the Phidata.

Doubao llm is created by from agno.models.openai.like import OpenAILike.

Error

openai.BadRequestError: Error code: 400 - {'error': {'code': 'InvalidParameter', 'message': 'One or more parameters specified in the request are not valid. Request id: 02173850780234684bae976ab37fc49f95f8882bb751870475bb0', 'param': '', 'type': 'BadRequest'}}

Steps to Reproduce

code

from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.hackernews import HackerNewsTools
from agno.tools.duckduckgo import DuckDuckGoTools
from agno.tools.newspaper4k import Newspaper4kTools
from LLMPool.LLMManager import LLMModelManager

hn_researcher = Agent(
    name="HackerNews Researcher",
    model=LLMModelManager().get_llm_model("doubao"),
    role="Gets top stories from hackernews.",
    tools=[HackerNewsTools()],
)

web_searcher = Agent(
    name="Web Searcher",
    model=LLMModelManager().get_llm_model("doubao"),
    role="Searches the web for information on a topic",
    tools=[DuckDuckGoTools()],
    add_datetime_to_instructions=True,
)

article_reader = Agent(
    name="Article Reader",
    model=LLMModelManager().get_llm_model("doubao"),
    role="Reads articles from URLs.",
    tools=[Newspaper4kTools()],
)

hn_team = Agent(
    name="Hackernews Team",
    model=LLMModelManager().get_llm_model("doubao"),
    team=[hn_researcher, web_searcher, article_reader],
    instructions=[
        "First, search hackernews for what the user is asking about.",
        "Then, ask the article reader to read the links for the stories to get more information.",
        "Important: you must provide the article reader with the links to read.",
        "Then, ask the web searcher to search for each story to get more information.",
        "Finally, provide a thoughtful and engaging summary.",
    ],
    show_tool_calls=True,
    markdown=True,
)
hn_team.print_response("Write an article about the top 2 stories on hackernews", stream=False)

[manthanguptaa]

Hey @JokerHB! Can you help with the steps on how you configured Doubao?

[JokerHB]

Hey @JokerHB! Can you help with the steps on how you configured Doubao?

Thanks a lot.

Here is the code for using Doubao to run the demo. Additionally, I added "introduce yourself" instructions for all agents and the agent team. For individual agents, it works well, but it fails for the agent team. The error message is:

openai.BadRequestError: Error code: 400 - {'error': {'code': 'InvalidParameter', 'message': 'One or more parameters specified in the request are not valid. Request id: 021738556852298066e25a9d76d456a17836e5a68876f93efddaa', 'param': '', 'type': 'BadRequest'}}

Code

# -*- coding: utf-8 -*-

import os

from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.tools.hackernews import HackerNewsTools
from agno.tools.duckduckgo import DuckDuckGoTools
from agno.tools.newspaper4k import Newspaper4kTools
from agno.models.openai.like import OpenAILike

hn_researcher = Agent(
    name="HackerNews Researcher",
    model=OpenAILike(id=os.getenv("DOUBAO_MODEL_ID"), base_url=os.getenv("DOUBAO_BASE_URL"), api_key=os.getenv("DOUBAO_API_KEY")),
    role="Gets top stories from hackernews.",
    tools=[HackerNewsTools()],
)

web_searcher = Agent(
    name="Web Searcher",
    model=OpenAILike(id=os.getenv("DOUBAO_MODEL_ID"), base_url=os.getenv("DOUBAO_BASE_URL"), api_key=os.getenv("DOUBAO_API_KEY")),
    role="Searches the web for information on a topic",
    tools=[DuckDuckGoTools()],
    add_datetime_to_instructions=True,
)

article_reader = Agent(
    name="Article Reader",
    model=OpenAILike(id=os.getenv("DOUBAO_MODEL_ID"), base_url=os.getenv("DOUBAO_BASE_URL"), api_key=os.getenv("DOUBAO_API_KEY")),
    role="Reads articles from URLs.",
    tools=[Newspaper4kTools()],
)

hn_team = Agent(
    name="Hackernews Team",
    model=OpenAILike(id=os.getenv("DOUBAO_MODEL_ID"), base_url=os.getenv("DOUBAO_BASE_URL"), api_key=os.getenv("DOUBAO_API_KEY")),
    team=[hn_researcher, web_searcher, article_reader],
    instructions=[
        "First, search hackernews for what the user is asking about.",
        "Then, ask the article reader to read the links for the stories to get more information.",
        "Important: you must provide the article reader with the links to read.",
        "Then, ask the web searcher to search for each story to get more information.",
        "Finally, provide a thoughtful and engaging summary.",
    ],
    show_tool_calls=True,
    markdown=True,
)


hn_researcher.print_response("Please introduce yourself", stream=False)
web_searcher.print_response("Please introduce yourself", stream=False)
article_reader.print_response("Please introduce yourself", stream=False)
hn_team.print_response("Please introduce yourself", stream=False)

Here is the request.content for all agents and the agent team.

hn_researcher

b'{"messages": [{"role": "system", "content": "<your_role>\\nGets top stories from hackernews.\\n</your_role>"}, {"role": "user", "content": "Please introduce yourself"}], "model": "ep-20250103180010-kv47b", "tool_choice": "auto", "tools": [{"type": "function", "function": {"name": "get_top_hackernews_stories",
"description": "Use this function to get top stories from Hacker News.", "parameters": {"type": "object", "properties": {"num_stories": {"type": "number", "description": "(int) Number of stories to return. Defaults to 10."}}, "required": []}}}, {"type": "function", "function": {"name": "get_user_details", 
"description": "Use this function to get the details of a Hacker News user using their username.", "parameters": {"type": "object", "properties": {"username": {"type": "string", "description": "(str) Username of the user to get details for."}}, "required": ["username"]}}}]}'

web_searcher

b'{"messages": [{"role": "system", "content": "<your_role>\\nSearches the web for information on a topic\\n</your_role>\\n\\n<additional_information>\\n- The current time is 2025-02-03 19:40:23.418357\\n</additional_information>"}, {"role": "user", "content": "Please introduce yourself"}], "model": 
"ep-20250103180010-kv47b", "tool_choice": "auto", "tools": [{"type": "function", "function": {"name": "duckduckgo_search", "description": "Use this function to search DuckDuckGo for a query.", "parameters": {"type": "object", "properties": {"query": {"type": "string", "description": "(str) The query to search 
for."}, "max_results": {"type": "number", "description": "(optional, default=5) The maximum number of results to return."}}, "required": ["query"]}}}, {"type": "function", "function": {"name": "duckduckgo_news", "description": "Use this function to get the latest news from DuckDuckGo.", "parameters": {"type": 
"object", "properties": {"query": {"type": "string", "description": "(str) The query to search for."}, "max_results": {"type": "number", "description": "(optional, default=5) The maximum number of results to return."}}, "required": ["query"]}}}]}'

article_reader

b'{"messages": [{"role": "system", "content": "<your_role>\\nReads articles from URLs.\\n</your_role>"}, {"role": "user", "content": "Please introduce yourself"}], "model": "ep-20250103180010-kv47b", "tool_choice": "auto", "tools": [{"type": "function", "function": {"name": "read_article", "description": "Use 
this function to read an article from a URL.", "parameters": {"type": "object", "properties": {"url": {"type": "string", "description": "(str) The URL of the article."}}, "required": ["url"]}}}]}'

hn_team

b'{"messages": [{"role": "system", "content": "<agent_team>\\nYou are the leader of a team of AI Agents:\\n- You can either respond directly or transfer tasks to other Agents in your team depending on the tools available to them.\\n- If you transfer a task to another Agent, make sure to include:\\n  - 
task_description (str): A clear description of the task.\\n  - expected_output (str): The expected output.\\n  - additional_information (str): Additional information that will help the Agent complete the task.\\n- You must always validate the output of the other Agents before responding to the user.\\n- You can
re-assign the task if you are not satisfied with the result.\\n</agent_team>\\n\\n<instructions>\\n- First, search hackernews for what the user is asking about.\\n- Then, ask the article reader to read the links for the stories to get more information.\\n- Important: you must provide the article reader with the
links to read.\\n- Then, ask the web searcher to search for each story to get more information.\\n- Finally, provide a thoughtful and engaging summary.\\n</instructions>\\n\\n<additional_information>\\n- Use markdown to format your answers.\\n</additional_information>\\n\\n<transfer_instructions>\\nYou can 
transfer tasks to the following Agents in your team:\\n\\nAgent 1:\\nName: HackerNews Researcher\\nRole: Gets top stories from hackernews.\\nAvailable tools: get_top_hackernews_stories, get_user_details\\n\\nAgent 2:\\nName: Web Searcher\\nRole: Searches the web for information on a topic\\nAvailable tools: 
duckduckgo_search, duckduckgo_news\\n\\nAgent 3:\\nName: Article Reader\\nRole: Reads articles from URLs.\\nAvailable tools: read_article\\n</transfer_instructions>"}, {"role": "user", "content": "Please introduce yourself"}], "model": "ep-20250103180010-kv47b", "tool_choice": "auto", "tools": [{"type": 
"function", "function": {"name": "transfer_task_to_hackernews_researcher", "description": "Use this function to transfer a task to hackernews_researcher\\nYou must provide a clear and concise description of the task the agent should achieve AND the expected output.\\nArgs:\\n    task_description (str): A clear 
and concise description of the task the agent should achieve.\\n    expected_output (str): The expected output from the agent.\\n    additional_information (Optional[str]): Additional information that will help the agent complete the task.\\nReturns:\\n    str: The result of the delegated task.\\n", 
"parameters": {"type": "object", "properties": {"task_description": {"type": "string"}, "expected_output": {"type": "string"}, "additional_information": {"type": ["string", "null"]}}, "required": ["task_description", "expected_output"]}}}, {"type": "function", "function": {"name": 
"transfer_task_to_web_searcher", "description": "Use this function to transfer a task to web_searcher\\nYou must provide a clear and concise description of the task the agent should achieve AND the expected output.\\nArgs:\\n    task_description (str): A clear and concise description of the task the agent 
should achieve.\\n    expected_output (str): The expected output from the agent.\\n    additional_information (Optional[str]): Additional information that will help the agent complete the task.\\nReturns:\\n    str: The result of the delegated task.\\n", "parameters": {"type": "object", "properties": 
{"task_description": {"type": "string"}, "expected_output": {"type": "string"}, "additional_information": {"type": ["string", "null"]}}, "required": ["task_description", "expected_output"]}}}, {"type": "function", "function": {"name": "transfer_task_to_article_reader", "description": "Use this function to 
transfer a task to article_reader\\nYou must provide a clear and concise description of the task the agent should achieve AND the expected output.\\nArgs:\\n    task_description (str): A clear and concise description of the task the agent should achieve.\\n    expected_output (str): The expected output from the
agent.\\n    additional_information (Optional[str]): Additional information that will help the agent complete the task.\\nReturns:\\n    str: The result of the delegated task.\\n", "parameters": {"type": "object", "properties": {"task_description": {"type": "string"}, "expected_output": {"type": "string"}, 
"additional_information": {"type": ["string", "null"]}}, "required": ["task_description", "expected_output"]}}}]}'