Update app.py

app.py

@@ -3,32 +3,161 @@ import gradio as gr
 import requests
 import inspect
 import pandas as pd
+from smolagents import OpenAIServerModel
+from smolagents import CodeAgent, Tool, tool
+from smolagents import DuckDuckGoSearchTool, VisitWebpageTool
+from smolagents import PythonInterpreterTool # Import the built-in Python Interpreter Tool
 
-# (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 ------
-class BasicAgent:
+# --- Tool Definitions ---
+
+class GaiaFileTool(Tool):
+    """
+    A smolagents.Tool subclass for downloading files from the GAIA API.
+    """
+    name = "download_gaia_file"
+    description = "Downloads a file associated with a given GAIA task ID and returns its content. It takes 'task_id' as input and returns the file content as a string. Use this when a question refers to an external file."
+    inputs = {"task_id": {"type": "str", "description": "The task ID for which to download the file (e.g., '2345')."}}
+    output_type = str
+
+    def __init__(self, api_base_url=DEFAULT_API_URL):
+        super().__init__()
+        self.api_base_url = api_base_url
+        print(f"GaiaFileTool initialized with API base URL: {self.api_base_url}")
+
+    def forward(self, task_id: str) -> str:
+        """
+        The core logic for the tool: downloads a file from the GAIA API.
+        This method is called by the agent when it uses this tool.
+        """
+        file_url = f"{self.api_base_url}/files/{task_id}"
+        print(f"Attempting to download file from: {file_url}")
+        try:
+            response = requests.get(file_url)
+            response.raise_for_status()
+            print(f"Successfully downloaded file for task_id {task_id}")
+            return response.text
+        except requests.exceptions.RequestException as e:
+            print(f"Error downloading file for task_id {task_id}: {e}")
+            return f"Error downloading file: {e}"
+
+# Removed the custom python_repl function as we are using the built-in tool
+
+# --- Custom GAIA Agent Definition ---
+class GaiaAgent(CodeAgent):
+    """
+    A smolagents-based agent designed to tackle GAIA Level 1 benchmark questions.
+    It uses Gemini Flash for reasoning and integrates a Python Interpreter, a
+    GAIA file download tool, and web browsing/searching tools.
+    """
     def __init__(self):
-        print("BasicAgent initialized.")
+        print("GaiaAgent initializing...")
+        gemini_api_key = os.getenv("GEMINI_API_KEY")
+        if not gemini_api_key:
+            print("WARNING: GEMINI_API_KEY environment variable not set.")
+            print("Please set GEMINI_API_KEY for Gemini Flash to work.")
+
+        self.llm_model = OpenAIServerModel(
+            model_id="gemini-2.0-flash",
+            api_base="https://generativelanguage.googleapis.com/v1beta/openai/",
+            api_key=gemini_api_key,
+            temperature=0.1,
+        )
+
+        # Initialize GAIA file tool
+        gaia_file_tool_instance = GaiaFileTool()
+
+        # Initialize web searching and browsing tools
+        duckduckgo_search_tool = DuckDuckGoSearchTool()
+        visit_webpage_tool = VisitWebpageTool()
+
+        # Initialize the built-in Python Interpreter Tool
+        # By default, PythonInterpreterTool uses a local executor, which is generally safe
+        # for controlled environments like Hugging Face Spaces.
+        python_interpreter_tool = PythonInterpreterTool()
+
+        # Define the tools available to the agent
+        agent_tools = [
+            python_interpreter_tool, # Use the built-in Python interpreter tool
+            gaia_file_tool_instance,
+            duckduckgo_search_tool,
+            visit_webpage_tool
+        ]
+        super().__init__(model=self.llm_model, tools=agent_tools)
+        print("GaiaAgent initialized successfully with Gemini Flash and built-in tools.")
+
     def __call__(self, question: str) -> str:
-        print(f"Agent received question (first 50 chars): {question[:50]}...")
-        fixed_answer = "This is a default answer."
-        print(f"Agent returning fixed answer: {fixed_answer}")
-        return fixed_answer
+        """
+        The main method for the agent to process a question and return an answer.
+        This will involve the agent's internal reasoning, tool use, and planning.
+        """
+        print(f"\n--- Agent received question (first 100 chars): {question[:100]}...")
 
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+        try:
+            prompt = (
+                f"You are an AI agent designed to solve GAIA benchmark questions. "
+                f"Your goal is to provide the exact answer as a string, without any additional text, "
+                f"explanation, or the phrase 'FINAL ANSWER:'. "
+                f"Break down the problem, use the available tools (python_interpreter, download_gaia_file, "
+                f"duckduckgo_search_tool, visit_webpage_tool) as needed, and think step-by-step. "
+                f"Use 'python_interpreter' for any calculations or code execution. "
+                f"Use 'duckduckgo_search_tool' to find information on the web. "
+                f"Use 'visit_webpage_tool' to read the content of a specific URL. "
+                f"When you have the final answer, output ONLY the answer string.\n\n"
+                f"Question: {question}"
+            )
+            
+            result = self.run(query=prompt)
+
+            final_answer = self._extract_exact_answer(result)
+            print(f"--- Agent returning final answer (first 100 chars): {final_answer[:100]}...")
+            return final_answer
+        except Exception as e:
+            print(f"--- Error during agent execution: {e}")
+            return "Agent encountered an error and could not provide an answer."
+
+    def _extract_exact_answer(self, raw_output: str) -> str:
+        """
+        Extracts and formats the exact answer from the agent's raw output.
+        Ensures no "FINAL ANSWER" text is included and handles any
+        extraneous formatting. This function is crucial for GAIA's exact match scoring.
+        """
+        cleaned_output = raw_output.replace("FINAL ANSWER:", "").strip()
+        cleaned_output = cleaned_output.replace("Answer:", "").strip()
+        cleaned_output = cleaned_output.replace("The answer is:", "").strip()
+        cleaned_output = cleaned_output.replace("```python", "").replace("```", "").strip()
+
+        lines = cleaned_output.split('\n')
+        if lines:
+            potential_answer = lines[-1].strip()
+            if len(potential_answer) < 5 or "tool_code" in potential_answer.lower():
+                for line in reversed(lines[:-1]):
+                    if line.strip() and "tool_code" not in line.lower():
+                        potential_answer = line.strip()
+                        break
+            cleaned_output = potential_answer
+
+        if cleaned_output.startswith('"') and cleaned_output.endswith('"'):
+            cleaned_output = cleaned_output[1:-1]
+        if cleaned_output.startswith("'") and cleaned_output.endswith("'"):
+            cleaned_output = cleaned_output[1:-1]
+
+        return cleaned_output.strip()
+
+
+# --- Gradio Application Logic (mostly unchanged from template) ---
+
+def run_and_submit_all(profile: gr.OAuthProfile | None):
     """
-    Fetches all questions, runs the BasicAgent on them, submits all answers,
+    Fetches all questions, runs the GaiaAgent on them, submits all answers,
     and displays the results.
     """
-    # --- Determine HF Space Runtime URL and Repo URL ---
-    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
+    space_id = os.getenv("SPACE_ID")
 
     if profile:
-        username= f"{profile.username}"
+        username = f"{profile.username}"
         print(f"User logged in: {username}")
     else:
         print("User not logged in.")
@@ -38,123 +167,70 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
     questions_url = f"{api_url}/questions"
     submit_url = f"{api_url}/submit"
 
-    # 1. Instantiate Agent ( modify this part to create your agent)
     try:
-        agent = BasicAgent()
+        agent = GaiaAgent()
     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 ( usefull for others so please keep it public)
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(agent_code)
 
-    # 2. Fetch Questions
-    print(f"Fetching questions from: {questions_url}")
     try:
-        response = requests.get(questions_url, timeout=15)
-        response.raise_for_status()
-        questions_data = response.json()
-        if not questions_data:
-             print("Fetched questions list is empty.")
-             return "Fetched questions list is empty or invalid format.", None
-        print(f"Fetched {len(questions_data)} questions.")
+        print(f"Fetching questions from: {questions_url}")
+        questions_response = requests.get(questions_url)
+        questions_response.raise_for_status()
+        questions = questions_response.json()
+        print(f"Fetched {len(questions)} questions.")
     except requests.exceptions.RequestException as e:
-        print(f"Error fetching questions: {e}")
         return f"Error fetching questions: {e}", None
-    except requests.exceptions.JSONDecodeError as e:
-         print(f"Error decoding JSON response from questions endpoint: {e}")
-         print(f"Response text: {response.text[:500]}")
-         return f"Error decoding server response for questions: {e}", None
-    except Exception as e:
-        print(f"An unexpected error occurred fetching questions: {e}")
-        return f"An unexpected error occurred fetching questions: {e}", None
-
-    # 3. Run your Agent
-    results_log = []
-    answers_payload = []
-    print(f"Running agent on {len(questions_data)} questions...")
-    for item in questions_data:
-        task_id = item.get("task_id")
-        question_text = item.get("question")
-        if not task_id or question_text is None:
-            print(f"Skipping item with missing task_id or question: {item}")
-            continue
-        try:
-            submitted_answer = agent(question_text)
-            answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
-            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
-        except Exception as e:
-             print(f"Error running agent on task {task_id}: {e}")
-             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})
 
-    if not answers_payload:
-        print("Agent did not produce any answers to submit.")
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
+    all_answers = []
+    results_data = []
+
+    for i, q_data in enumerate(questions):
+        task_id = q_data.get("task_id", f"unknown_{i}")
+        question_text = q_data.get("question", "No question text found.")
+        print(f"\n--- Processing Task ID: {task_id} ---")
+        print(f"Question: {question_text[:100]}...")
 
-    # 4. Prepare Submission 
-    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
-    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
-    print(status_update)
+        agent_answer = agent(question_text)
+        all_answers.append({"task_id": task_id, "answer": agent_answer})
+        results_data.append({
+            "Task ID": task_id,
+            "Question": question_text,
+            "Agent Answer": agent_answer
+        })
+        print(f"--- Finished processing Task ID: {task_id} ---")
 
-    # 5. Submit
-    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
     try:
-        response = requests.post(submit_url, json=submission_data, timeout=60)
-        response.raise_for_status()
-        result_data = response.json()
-        final_status = (
-            f"Submission Successful!\n"
-            f"User: {result_data.get('username')}\n"
-            f"Overall Score: {result_data.get('score', 'N/A')}% "
-            f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
-            f"Message: {result_data.get('message', 'No message received.')}"
-        )
-        print("Submission successful.")
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
-    except requests.exceptions.HTTPError as e:
-        error_detail = f"Server responded with status {e.response.status_code}."
-        try:
-            error_json = e.response.json()
-            error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
-        except requests.exceptions.JSONDecodeError:
-            error_detail += f" Response: {e.response.text[:500]}"
-        status_message = f"Submission Failed: {error_detail}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
-    except requests.exceptions.Timeout:
-        status_message = "Submission Failed: The request timed out."
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        print(f"\nSubmitting {len(all_answers)} answers to: {submit_url}")
+        submission_payload = {
+            "username": username,
+            "code_link": f"https://huggingface.co/spaces/{space_id}/tree/main" if space_id else "local_execution",
+            "answers": all_answers
+        }
+        submit_response = requests.post(submit_url, json=submission_payload)
+        submit_response.raise_for_status()
+        submission_result = submit_response.json()
+        print(f"Submission successful: {submission_result}")
+        status_message = f"Submission successful!\nScore: {submission_result.get('score', 'N/A')}\nDetails: {submission_result.get('message', 'No message')}"
     except requests.exceptions.RequestException as e:
-        status_message = f"Submission Failed: Network error - {e}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
-    except Exception as e:
-        status_message = f"An unexpected error occurred during submission: {e}"
-        print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        print(f"Error submitting answers: {e}")
+        status_message = f"Error submitting answers: {e}"
 
+    results_df = pd.DataFrame(results_data)
+    return status_message, results_df
 
-# --- Build Gradio Interface using Blocks ---
+# --- Gradio UI ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
     gr.Markdown(
         """
-        **Instructions:**
+        # GAIA Level 1 Agent Evaluation
+        This application allows you to run your `smolagents`-based agent on the GAIA Level 1 benchmark
+        and submit your answers to the leaderboard.
 
-        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.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
-
-        ---
-        **Disclaimers:**
-        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
-        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
+        **Important:**
+        1. **Login to Hugging Face** using the button below to submit your score.
+        2. **Set `GEMINI_API_KEY`**: Ensure your `GEMINI_API_KEY` is set as a Space Secret
+           in Hugging Face Spaces (or as an environment variable if running locally)
+           for the Gemini Flash model to function.
         """
     )
 
@@ -163,7 +239,6 @@ with gr.Blocks() as demo:
     run_button = gr.Button("Run Evaluation & Submit All Answers")
 
     status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    # Removed max_rows=10 from DataFrame constructor
     results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
 
     run_button.click(
@@ -173,9 +248,8 @@ with gr.Blocks() as demo:
 
 if __name__ == "__main__":
     print("\n" + "-"*30 + " App Starting " + "-"*30)
-    # Check for SPACE_HOST and SPACE_ID at startup for information
     space_host_startup = os.getenv("SPACE_HOST")
-    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup
+    space_id_startup = os.getenv("SPACE_ID")
 
     if space_host_startup:
         print(f"✅ SPACE_HOST found: {space_host_startup}")
@@ -183,14 +257,11 @@ if __name__ == "__main__":
     else:
         print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
 
-    if space_id_startup: # Print repo URLs if SPACE_ID is found
+    if space_id_startup:
         print(f"✅ SPACE_ID found: {space_id_startup}")
         print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
         print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
     else:
-        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")
-
-    print("-"*(60 + len(" App Starting ")) + "\n")
+        print("ℹ️  SPACE_ID environment variable not found. Code link might be incorrect for submission.")
 
-    print("Launching Gradio Interface for Basic Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+    demo.launch()