Developed the basis

(cherry picked from commit 2a41ac40514a56388d6366704838754986f53aa2)

app.py

@@ -1,9 +1,8 @@
-from smolagents import CodeAgent,DuckDuckGoSearchTool, HfApiModel,load_tool,tool
+from smolagents import CodeAgent, HfApiModel,load_tool,tool
 import datetime
-import requests
 import pytz
 import yaml
-from tools.final_answer import FinalAnswerTool
+from src.tools.final_answer import FinalAnswerTool
 
 from Gradio_UI import GradioUI
 

requirements.txt

@@ -1,5 +1,15 @@
-markdownify
-smolagents
-requests
-duckduckgo_search
+markdownify~=1.1.0
+smolagents~=1.10.0
+requests~=2.32.3
+duckduckgo_search~=7.5.1
 pandas
+smolagents[litellm]
+matplotlib
+geopandas
+shapely
+kaleido
+pytz~=2025.1
+PyYAML~=6.0.2
+langchain~=0.3.20
+langchain-community~=0.3.19
+python-dotenv~=1.0.1

src/agent.py

@@ -0,0 +1,84 @@
+from smolagents import CodeAgent, GoogleSearchTool, VisitWebpageTool, HfApiModel
+
+from src.party_planner.tools.travel_time import calculate_cargo_travel_time
+
+
+def create_agent(
+        agent_type: str,
+        name: str,
+        model: HfApiModel,
+        tools: list,
+        max_steps: int = 10,
+        additional_imports: list[str] = None,
+        description: str = "",
+        interval: int = 0,
+        verbosity: int = 0,
+        **kwargs
+):
+    """
+    **kwargs can be: managed_agents, final_answer_checks, ...
+    """
+    if agent_type == "code_agent":
+        return CodeAgent(
+            name=name,
+            model=model,
+            tools=tools,
+            additional_authorized_imports=additional_imports,
+            max_steps=max_steps,
+            description=description,
+            planning_interval=interval,
+            verbosity_level=verbosity,
+            **kwargs
+        )
+    return None
+
+
+if __name__ == "__main__":
+    import os
+    from dotenv import load_dotenv
+    from src.model import get_model
+
+    load_dotenv()
+    # os.environ["SERPER_API_KEY"] = os.getenv("SERPER_API_KEY")
+    os.environ["SERPAPI_API_KEY"] = os.getenv("SERPAPI_API_KEY")
+
+    AgentType = "code_agent"
+    Name = "web_agent"
+    Model = get_model(
+        model_id="Qwen/Qwen2.5-Coder-32B-Instruct",
+        provider="hf-inference"  # "together"
+    )
+    ToolNames = [
+        GoogleSearchTool(provider="serpapi"),
+        VisitWebpageTool(),
+        calculate_cargo_travel_time
+    ]
+    AdditionalImports = ["pandas"]
+    MaxSteps = 3
+    Description = "Browses the web to find information"
+    Interval = 4
+
+    # Simple agent served as a baseline for the multi-agent system
+    Agent = create_agent(
+        agent_type=AgentType,
+        name=Name,
+        model=Model,
+        tools=ToolNames,
+        additional_imports=AdditionalImports,
+        max_steps=MaxSteps,
+        description=Description,
+        interval=Interval
+    )
+
+    Task = """Find all Batman filming locations in the world, calculate the time to transfer via cargo plane to 
+    here (we're in Gotham, 40.7128° N, 74.0060° W), and return them to me as a pandas dataframe. Also give me some 
+    supercar factories with the same cargo plane transfer time."""
+    Prompt = f"""
+    You're an expert analyst. You make comprehensive reports after visiting many websites.
+    Don't hesitate to search for many queries at once in a for loop.
+    For each data point that you find, visit the source url to confirm numbers.
+
+    {Task}
+    """
+    result = Agent.run(Prompt)
+    print('\n' * 2, result)

src/party_planner/utils.py

@@ -0,0 +1,40 @@
+from smolagents.utils import encode_image_base64, make_image_url
+from smolagents import OpenAIServerModel
+from PIL import Image
+
+
+def check_reasoning_and_plot(
+        final_answer,
+        agent_memory,
+        map_image_path: str = "src/party_planner/saved_map.png"
+) -> bool:
+    final_answer
+    multimodal_model = OpenAIServerModel("gpt-4o", max_tokens=2048)
+    image = Image.open(map_image_path)
+    prompt = (
+        f"Here is a user-given task and the agent steps: {agent_memory.get_succinct_steps()}. Now here is the plot that was made."
+        "Please check that the reasoning process and plot are correct: do they correctly answer the given task?"
+        "First list reasons why yes/no, then write your final decision: PASS in caps lock if it is satisfactory, FAIL if it is not."
+        "Don't be harsh: if the plot mostly solves the task, it should pass."
+        "To pass, a plot should be made using px.scatter_map and not any other method (scatter_map looks nicer)."
+    )
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {
+                    "type": "text",
+                    "text": prompt,
+                },
+                {
+                    "type": "image_url",
+                    "image_url": {"url": make_image_url(encode_image_base64(image))},
+                },
+            ],
+        }
+    ]
+    output = multimodal_model(messages).content
+    print("Feedback: ", output)
+    if "FAIL" in output:
+        raise Exception(output)
+    return True

{tools → src/tools}/visit_webpage.py

@@ -3,6 +3,7 @@ from smolagents.tools import Tool
 import requests
 import markdownify
 import smolagents
+import re
 
 class VisitWebpageTool(Tool):
     name = "visit_webpage"