Refactor SmolAgent to integrate OpenAI's API, enhancing question answering capabilities with improved instructions and error handling. Update Gradio interface for GAIA evaluation submission and results display.

app.py

@@ -2,50 +2,60 @@ import os
 import gradio as gr
 import requests
 import pandas as pd
-from smolagents import CodeAgent, InferenceClientModel, DuckDuckGoSearchTool
 from dotenv import load_dotenv
 
-# Load environment variables
+from smolagents import CodeAgent, OpenAIServerModel, DuckDuckGoSearchTool
+
+# Load environment variables (including OPENAI_API_KEY)
 load_dotenv()
 
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
-HF_TOKEN = os.getenv("HF_TOKEN")
-INSTRUCTIONS = """Your task is to answer the user's question.
-Return a number OR as few words as possible OR a comma separated list of numbers and/or strings.
-- If you are asked for a number, don't use a comma to write your number, and don't use units like $ or % unless specified otherwise.
-- If you are asked for a string, don't use articles or abbreviations (e.g., for cities), and write digits in plain text unless specified otherwise.
-- If you are asked for a comma-separated list, apply the above rules to each element.
-"""
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+
+INSTRUCTIONS = """You are a general AI assistant. I will ask you a question. Report your thoughts, and then provide your final answer.
+
+CRITICAL FORMATTING RULES:
+- Your final answer should be a number OR as few words as possible OR a comma separated list of numbers and/or strings
+- If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise
+- If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise
+- If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string
+- Be extremely precise with spelling and formatting - the evaluation uses exact matching
+- For strings: no extra spaces, no punctuation unless part of the answer, lowercase
+- For numbers: just the number, no units, no commas, no currency symbols
+- Provide ONLY the answer as your final response, nothing else
+
+You have access to a web search tool to help you find accurate information. Use it when you need to look up facts."""
 
 # --- Smol Agent Definition ---
 class SmolAgent:
-    def __init__(self, hf_token: str):
-        print("Initializing SmolAgent with smolagents...")
-        if not hf_token:
-            raise ValueError("Hugging Face token not found. Please set HF_TOKEN environment variable.")
-
-        # Initialize the model
-        model = InferenceClientModel(
-            model_id="meta-llama/Meta-Llama-3-8B-Instruct",
-            token=hf_token,
+    def __init__(self):
+        print("Initializing SmolAgent with OpenAI model...")
+        if not OPENAI_API_KEY:
+            raise ValueError("OPENAI_API_KEY not found. Please set it in your environment.")
+
+        # Initialize the OpenAI-backed model
+        self.model = OpenAIServerModel(
+            model_id="gpt-4o-mini",                  # or "gpt-4", "gpt-3.5-turbo", etc.
+            api_base="https://api.openai.com/v1",
+            api_key=OPENAI_API_KEY,
         )
 
         # Initialize the agent with tools and instructions
         self.agent = CodeAgent(
             tools=[DuckDuckGoSearchTool()],
-            model=model,
+            model=self.model,
             instructions=INSTRUCTIONS,
+            max_steps=7,
         )
         print("SmolAgent initialized with CodeAgent and DuckDuckGoSearchTool.")
 
     def __call__(self, question: str) -> str:
         print(f"\n🪐 Running on question:\n{question}\n")
         try:
-            # The CodeAgent's run method returns the final answer directly
             answer = self.agent.run(question)
             print(f"✅ Agent's final answer: {answer}")
-            return str(answer) # Ensure the output is a string
+            return str(answer)
         except Exception as e:
             import traceback
             traceback.print_exc()
@@ -53,184 +63,145 @@ class SmolAgent:
             print(f"❌ {error_message}")
             return error_message
 
-
-def run_and_submit_all(profile: gr.OAuthProfile | None):
-    """
-    Fetches all questions, runs the SmolAgent on them, submits all answers,
-    and displays the results.
-    """
-    # --- Determine HF Space Runtime URL and Repo URL ---
-    space_id = os.getenv("SPACE_ID")
-
-    if profile:
-        username = f"{profile.username}"
-        print(f"User logged in: {username}")
-    else:
-        print("User not logged in.")
-        return "Please Login to Hugging Face with the button.", None
-
-    api_url = DEFAULT_API_URL
-    questions_url = f"{api_url}/questions"
-    submit_url = f"{api_url}/submit"
-
-    # 1. Instantiate Agent
+def run_gaia_evaluation(username: str):
+    """Run the complete GAIA evaluation and submit results"""
+    print("🚀 GAIA Benchmark Evaluation with ChatGPT")
+    print("=" * 60)
+    
+    if not username:
+        return "❌ Please provide a username"
+    
+    print(f"👤 User: {username}")
+    
+    # Initialize the agent
     try:
-        agent = SmolAgent(hf_token=HF_TOKEN)
+        agent = SmolAgent()
     except Exception as e:
-        print(f"Error instantiating agent: {e}")
-        return f"Error initializing agent: {e}", None
-    
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(f"Agent code link: {agent_code}")
+        return f"❌ Failed to initialize agent: {e}"
 
-    # 2. Fetch Questions
-    print(f"Fetching questions from: {questions_url}")
+    # Fetch questions
     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.")
-    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
+        resp = requests.get(f"{DEFAULT_API_URL}/questions", timeout=30)
+        resp.raise_for_status()
+        data = resp.json()
+        questions = data if isinstance(data, list) else data.get("questions", [])
+        print(f"📋 Loaded {len(questions)} questions")
+    except requests.RequestException as e:
+        return f"❌ Error fetching questions: {e}"
+
+    # Process questions
+    results = []
+    progress_log = []
+    
+    for i, q in enumerate(questions):
+        task_id = q["task_id"]
+        text = q["question"]
+        progress_log.append(f"❓ Question {i+1}: {text}")
+        print(f"\n❓ Question {i+1}: {text}")
+
         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}"})
+            result = agent(text)
+            result_str = str(result).strip()
 
-    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)
+            # Take the last line as the answer
+            out = result_str.splitlines()[-1] if result_str else "AGENT ERROR: No response."
 
-    # 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)
+            if out.startswith("{"):
+                out = "AGENT ERROR: No final answer."
 
-    # 5. Submit
-    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
+            out = out.strip().rstrip(".")
+            results.append({"task_id": task_id, "submitted_answer": out})
+            
+            progress_log.append(f"✅ Answer: '{out}'")
+            print(f"✅ Answer: '{out}'")
+            
+        except Exception as e:
+            error_msg = f"AGENT ERROR: {e}"
+            results.append({"task_id": task_id, "submitted_answer": error_msg})
+            progress_log.append(f"❌ Error: {error_msg}")
+            print(f"❌ Error: {error_msg}")
+
+    # Submit results
+    payload = {
+        "username": username,
+        "agent_code": "chatgpt-gpt4o-mini-with-tools",
+        "answers": results,
+    }
+    
     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("📤 Submitting to GAIA leaderboard...")
+        post = requests.post(f"{DEFAULT_API_URL}/submit", json=payload, timeout=60)
+        post.raise_for_status()
+        res = post.json()
+        
+        # Format results for display
+        result_summary = f"""
+🏆 GAIA BENCHMARK RESULTS
+{'=' * 60}
+👤 User: {res.get('username', username)}
+📊 Overall Score: {res.get('score', res.get('overall_score', 'N/A'))}%
+✅ Correct: {res.get('correct_count', res.get('num_correct', 'N/A'))}/{len(results)}
+💬 Message: {res.get('message', 'N/A')}
+{'=' * 60}
+"""
+        
+        # Combine progress log with final results
+        full_log = "\n".join(progress_log) + "\n" + result_summary
+        return full_log
+        
+    except requests.RequestException as e:
+        error_msg = f"❌ Error submitting: {e}"
+        done = sum(1 for r in results if not r["submitted_answer"].startswith("AGENT ERROR"))
+        local_summary = f"📋 Completed locally: {done}/{len(results)}"
+        return "\n".join(progress_log) + "\n" + error_msg + "\n" + local_summary
+
+# --- Gradio Interface ---
+def create_interface():
+    with gr.Blocks(title="GAIA Benchmark with ChatGPT", theme=gr.themes.Soft()) as demo:
+        gr.Markdown("# 🚀 GAIA Benchmark Evaluation with ChatGPT")
+        gr.Markdown("This app runs the GAIA benchmark using ChatGPT (GPT-4o-mini) with web search capabilities.")
+        
+        with gr.Row():
+            with gr.Column(scale=1):
+                username_input = gr.Textbox(
+                    label="Hugging Face Username",
+                    placeholder="Enter your HF username",
+                    info="This will be used for the GAIA leaderboard submission"
+                )
+                
+                run_button = gr.Button("🚀 Run GAIA Evaluation", variant="primary", size="lg")
+                
+            with gr.Column(scale=2):
+                output_area = gr.Textbox(
+                    label="Results & Progress",
+                    lines=20,
+                    max_lines=50,
+                    interactive=False
+                )
+        
+        # Event handler
+        run_button.click(
+            fn=run_gaia_evaluation,
+            inputs=[username_input],
+            outputs=[output_area]
         )
-        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
-    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
-
-
-# --- Build Gradio Interface using Blocks ---
-with gr.Blocks() as demo:
-    gr.Markdown("# SmolLM Agent Evaluation Runner")
-    gr.Markdown(
-        """
-        **Instructions:**
-
-        1.  This space uses a `smolagents.CodeAgent` with the `meta-llama/Meta-Llama-3-8B-Instruct` model.
-        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.
-
-        ---
-        **Model Information:**
-        - Agent: `smolagents.CodeAgent`
-        - Model: `meta-llama/Meta-Llama-3-8B-Instruct`
-        - Tools: `DuckDuckGoSearchTool`
-
-        **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.
-        """
-    )
-
-    gr.LoginButton()
-
-    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(
-        fn=run_and_submit_all,
-        outputs=[status_output, results_table]
-    )
+        
+        gr.Markdown("""
+        ### How it works:
+        1. Enter your Hugging Face username
+        2. Click "Run GAIA Evaluation" 
+        3. The agent will process all 20 GAIA questions using ChatGPT + web search
+        4. Results will be automatically submitted to the GAIA leaderboard
+        5. Your score will be displayed here
+        
+        ### Requirements:
+        - Set `OPENAI_API_KEY` in your environment variables
+        - Valid Hugging Face username for leaderboard submission
+        """)
+    
+    return demo
 
+# --- Main execution ---
 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
-
-    if space_host_startup:
-        print(f"✅ SPACE_HOST found: {space_host_startup}")
-        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
-    else:
-        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
-
-    if space_id_startup: # Print repo URLs if SPACE_ID is found
-        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("Launching Gradio Interface for Smol Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+    demo = create_interface()
+    demo.launch()