Update app.py

app.py

@@ -1,116 +1,36 @@
-import base64
-import inspect
-import mimetypes
 import os
-import tempfile
-
 import gradio as gr
-import pandas as pd
 import requests
-from langchain_community.document_loaders import UnstructuredExcelLoader
-
-from agents import build_supervisor_agent
-
+import inspect
+import pandas as pd
+from smolagents import CodeAgent, DuckDuckGoSearchTool, OpenAIServerModel
 # (Keep Constants as is)
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
+
 # --- Basic Agent Definition ---
 # ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
-def fetch_file(task_id):
-    if not task_id:
-        return None
-    url = f"https://agents-course-unit4-scoring.hf.space/files/{task_id}"
-    response = requests.get(url)
-    if response.status_code == 200:
-        return response.content  # Return raw bytes
-    else:
-        print(f"Failed to fetch file for task_id {task_id}: {response.status_code}")
-        return None
-
-def build_multimodal_message(question, file_bytes=None, file_name=None):
-    """
-    Build a multimodal message with correct content blocks for text, image, audio, or file.
-    For .xlsx files, extract the text and append it to the question, since LLMs do not natively support .xlsx.
-    Follows: https://python.langchain.com/docs/how_to/multimodal_inputs/
-    """
-    content = []
-    # Special handling for .xlsx files
-    if file_bytes and file_name and file_name.lower().endswith('.xlsx'):
-        with tempfile.NamedTemporaryFile(delete=False, suffix='.xlsx') as tmp:
-            tmp.write(file_bytes)
-            tmp_path = tmp.name
-        loader = UnstructuredExcelLoader(tmp_path, mode="elements")
-        docs = loader.load()
-        excel_text = "\n".join(doc.page_content for doc in docs)
-        question = f"{question}\n\n[Excel file content follows:]\n{excel_text}"
-    content.append({"type": "text", "text": question})
-    if file_bytes and file_name and not file_name.lower().endswith('.xlsx'):
-        ext = file_name.lower().split('.')[-1]
-        b64_data = base64.b64encode(file_bytes).decode("utf-8")
-        mime_type, _ = mimetypes.guess_type(file_name)
-        # Handle common audio/image types explicitly
-        if ext in ["png"]:
-            mime_type = "image/png"
-            block_type = "image"
-        elif ext in ["jpg", "jpeg"]:
-            mime_type = "image/jpeg"
-            block_type = "image"
-        elif ext == "mp3":
-            mime_type = "audio/mpeg"
-            block_type = "audio"
-        elif ext == "wav":
-            mime_type = "audio/wav"
-            block_type = "audio"
-        elif ext == "m4a":
-            mime_type = "audio/mp4"
-            block_type = "audio"
-        else:
-            block_type = "file"
-            if not mime_type:
-                mime_type = "application/octet-stream"
-        block = {
-            "type": block_type,
-            "source_type": "base64",
-            "data": b64_data,
-            "mime_type": mime_type,
-            "filename": file_name,
-        }
-        content.append(block)
-    return [{"role": "user", "content": content}]
-
-def filter_supported_content_blocks(messages):
-    allowed_types = {"text", "image_url", "input_audio", "refusal", "audio", "file", "image"}
-    filtered = []
-    for msg in messages:
-        if "content" in msg and isinstance(msg["content"], list):
-            filtered_content = [block for block in msg["content"] if block.get("type") in allowed_types]
-            msg = dict(msg)
-            msg["content"] = filtered_content
-        filtered.append(msg)
-    return filtered
-
-class Agent:
-    def __init__(self, main_agent):
-        self.main_agent = main_agent
-        print("Agent initialized.")
-    def __call__(self, question: str, file_name: str = "", task_id: str = "") -> str:
+class BasicAgent:
+    def __init__(self):
+        print("BasicAgent initialized.")
+        # Initialize the model
+        #model = HfApiModel()
+        model = OpenAIServerModel(model_id="gpt-4.1")
+        # Initialize the search tool
+        search_tool = DuckDuckGoSearchTool()
+        # Initialize Agent
+        self.agent = CodeAgent(
+            model = model,
+            tools=[search_tool]
+        )
+    def __call__(self, question: str) -> str:
         print(f"Agent received question (first 50 chars): {question[:50]}...")
-        file_bytes = fetch_file(task_id) if file_name else None
-        message = build_multimodal_message(question, file_bytes, file_name)
-        # Filter out unsupported content block types
-        message = filter_supported_content_blocks(message)
-        result = self.main_agent.invoke({"messages": message})
-        answer = result["messages"][-1]
-        content = answer.content
-        if isinstance(content, list) and content and isinstance(content[0], dict) and "text" in content[0]:
-            return content[0]["text"]
-        elif isinstance(content, str):
-            return content
-        else:
-            return str(content)
+        fixed_answer =self.agent.run(question)
+        print(f"Agent returning fixed answer: {fixed_answer}")
+        return fixed_answer
 
-def run_and_submit_all(profile: gr.OAuthProfile | None, openai_key: str, google_key: str):
+def run_and_submit_all( profile: gr.OAuthProfile | None):
     """
     Fetches all questions, runs the BasicAgent on them, submits all answers,
     and displays the results.
@@ -119,7 +39,7 @@ def run_and_submit_all(profile: gr.OAuthProfile | None, openai_key: str, google_
     space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
 
     if profile:
-        username = f"{profile.username}"
+        username= f"{profile.username}"
         print(f"User logged in: {username}")
     else:
         print("User not logged in.")
@@ -131,11 +51,11 @@ def run_and_submit_all(profile: gr.OAuthProfile | None, openai_key: str, google_
 
     # 1. Instantiate Agent ( modify this part to create your agent)
     try:
-        agent = Agent(build_supervisor_agent(openai_key, google_key))
+        agent = BasicAgent()
     except Exception as e:
         print(f"Error instantiating agent: {e}")
         return f"Error initializing agent: {e}", None
-    # In the case of an app running as a hugging Face space, this link points toward your codebase ( useful for others so please keep it public)
+    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
     agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
     print(agent_code)
 
@@ -239,7 +159,7 @@ with gr.Blocks() as demo:
         **Instructions:**
         1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
         2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
-        3.  Enter your OpenAI and Google API keys below (if required by your agent).
+        3.  Enter your OpenAI key below (if required by your agent).
         4.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
         ---
         **Disclaimers:**
@@ -251,7 +171,6 @@ with gr.Blocks() as demo:
     gr.LoginButton()
 
     openai_key_box = gr.Textbox(label="OpenAI API Key", type="password", placeholder="sk-...", lines=1)
-    google_key_box = gr.Textbox(label="Google API Key", type="password", placeholder="AIza...", lines=1)
 
     run_button = gr.Button("Run Evaluation & Submit All Answers")
 
@@ -260,7 +179,7 @@ with gr.Blocks() as demo:
 
     run_button.click(
         fn=run_and_submit_all,
-        inputs=[openai_key_box, google_key_box],
+        inputs=[openai_key_box],
         outputs=[status_output, results_table]
     )