agents.py

from smolagents import OpenAIServerModel, CodeAgent, InferenceClientModel, DuckDuckGoSearchTool, VisitWebpageTool
import markdownify

import tools
import prompts

MANAGER_MODEL = "deepseek-ai/DeepSeek-R1"
AGENT_MODEL = "Qwen/Qwen2.5-Coder-32B-Instruct"
FINAL_ANSWER_MODEL = "deepseek-ai/DeepSeek-R1" # OpenAIServerModel
WEB_SEARCH_MODEL        = "Qwen/Qwen2.5-Coder-32B-Instruct"
IMAGE_ANALYSIS_MODEL    = "HuggingFaceM4/idefics2-8b"
AUDIO_ANALYSIS_MODEL    = "Qwen/Qwen2.5-Coder-32B-Instruct"
VIDEO_ANALYSIS_MODEL    = "Qwen/Qwen2.5-Coder-32B-Instruct"
YOUTUBE_ANALYSIS_MODEL  = "Qwen/Qwen2.5-Coder-32B-Instruct"
DOCUMENT_ANALYSIS_MODEL = "Qwen/Qwen2.5-Coder-32B-Instruct"
ARITHMETIC_MODEL        = "Qwen/Qwen2.5-Coder-32B-Instruct"
CODE_GENERATION_MODEL   = "Qwen/Qwen2.5-Coder-32B-Instruct"
CODE_EXECUTION_MODEL    = "Qwen/Qwen2.5-Coder-32B-Instruct"

# Agents

def create_custom_web_search_agent(message):
    return CodeAgent(
        name="custom_web_search_agent",
        description=prompts.get_web_search_prompt(message),
        model=InferenceClientModel(WEB_SEARCH_MODEL),
        max_steps=2,
        tools=[tools.simple_web_search_tool, tools.visit_web_page_tool],
    )

def create_simple_web_search_agent(message):
    return CodeAgent(
        name="simple_web_search_agent",
        description=prompts.get_web_search_prompt(message),
        model=InferenceClientModel(WEB_SEARCH_MODEL),
        max_steps=2,
        tools=[tools.simple_web_search_tool, tools.visit_web_page_tool],
    )

def create_image_analysis_agent(message):
    return CodeAgent(
        name="image_analysis_agent",
        description=prompts.get_image_analysis_prompt(message),
        model=InferenceClientModel(IMAGE_ANALYSIS_MODEL),
        tools=[tools.image_analysis_tool],
        max_steps=2,
    )

def create_audio_analysis_agent(message):
    return CodeAgent(
        name="audio_analysis_agent",
        description=prompts.get_audio_analysis_prompt(message),
        model=InferenceClientModel(AUDIO_ANALYSIS_MODEL),
        tools=[tools.audio_analysis_tool],
        max_steps=2,
    )

def create_manager_agent(message):
    simple_web_search_agent = create_simple_web_search_agent(message)
    image_analysis_agent = create_image_analysis_agent(message)
    audio_analysis_agent = create_audio_analysis_agent(message)

    return CodeAgent(
        name="manager_agent",
        model=InferenceClientModel(MANAGER_MODEL, provider="together", max_tokens=8096),
        description=prompts.get_manager_prompt(message),
        tools=[],
        planning_interval=4,
        verbosity_level=2,
        managed_agents=[
            simple_web_search_agent,
            image_analysis_agent,
            audio_analysis_agent,
        ],
        max_steps=10,
        additional_authorized_imports=[
            "requests",
            "zipfile",
            "os",
            "pandas",
            "numpy",
            "sympy",
            "json",
            "bs4",
            "pubchempy",
            "xml",
            "yahoo_finance",
            "Bio",
            "sklearn",
            "scipy",
            "pydub",
            "io",
            "PIL",
            "chess",
            "PyPDF2",
            "pptx",
            "torch",
            "datetime",
            "csv",
            "fractions",
        ],
    )

def create_final_answer_agent(message):
    return CodeAgent(
        name="final_answer_agent",
        description="Given a question and an initial answer, return the final refined answer following strict formatting rules.",
        model=InferenceClientModel(FINAL_ANSWER_MODEL),
        max_steps=3,
        tools=[],
    )