Update app.py

app.py

@@ -40,18 +40,135 @@ Question: "What is the capital?" -> Answer: "Paris"
 Question: "List the winners" → Answer: "John, Mary, Bob"
 """
 
+class SmolGaiaAgent:
+    """
+    Premium agent optimized for maximum accuracy on GAIA Level 1.
+    """
 
+    def __init__(self):
+        print("Initializing Premium SmolGaiaAgent...")
 
+        # Use the most capable model available
+        # Option 1: Qwen 32B (current - good balance)
+        self.model = OpenAIModel(
+            model_id="gpt-4.1",
+            api_key=os.getenv("OPENAI_API_KEY"),
+        )
 
+        # MORE STEPS = Better accuracy (but slower)
+        try:
+            self.agent = CodeAgent(
+                tools=[],
+                add_base_tools=True,
+                model=self.model,
+                max_steps=12,  # INCREASED from 6 to 12 for thorough reasoning
+                system_prompt=GAIA_SYSTEM_PROMPT,
+            )
+            print("Agent initialized with system_prompt parameter")
+            self.use_task_prefix = False
+        except TypeError as e:
+            print(f"system_prompt not supported, using task prefix: {e}")
+            self.agent = CodeAgent(
+                tools=[],
+                add_base_tools=True,
+                model=self.model,
+                max_steps=12,
+            )
+            self.use_task_prefix = True
+
+    def __call__(self, question: str) -> str:
+        """
+        Runs the CodeAgent on one question with enhanced answer extraction.
+        """
+        print(f"[Premium Agent] Question: {question[:80]}...")
+        
+        if self.use_task_prefix:
+            task = f"{GAIA_SYSTEM_PROMPT}\n\nTask: {question}"
+        else:
+            task = question
+            
+        try:
+            answer = self.agent.run(task)
+            answer = str(answer).strip()
+            
+            # Enhanced answer cleaning
+            answer = self.aggressive_clean_answer(answer)
+            
+            print(f"[Premium Agent] Final Answer: {answer}")
+            return answer
+        except Exception as e:
+            print(f"[Premium Agent] Error: {e}")
+            import traceback
+            traceback.print_exc()
+            return "Error processing question"
 
-
-
-
-# Submission logic Unchanged
+    def aggressive_clean_answer(self, answer: str) -> str:
+        """
+        Aggressively clean the answer to extract just the answer.
+        """
+        original = answer
+        
+        # Remove common prefixes (case insensitive)
+        prefixes_to_remove = [
+            "final answer:",
+            "the final answer is:",
+            "answer:",
+            "the answer is:",
+            "the answer is",
+            "result:",
+            "solution:",
+            "output:",
+        ]
+        
+        answer_lower = answer.lower()
+        for prefix in prefixes_to_remove:
+            if answer_lower.startswith(prefix):
+                answer = answer[len(prefix):].strip()
+                answer_lower = answer.lower()
+        
+        # Remove surrounding quotes
+        if (answer.startswith('"') and answer.endswith('"')) or \
+           (answer.startswith("'") and answer.endswith("'")):
+            answer = answer[1:-1].strip()
+        
+        # If answer contains "is:" extract what comes after
+        if " is:" in answer.lower():
+            parts = answer.split("is:")
+            if len(parts) > 1:
+                answer = parts[-1].strip()
+        
+        # If answer contains "are:" extract what comes after
+        if " are:" in answer.lower():
+            parts = answer.split("are:")
+            if len(parts) > 1:
+                answer = parts[-1].strip()
+        
+        # Remove trailing periods (unless it's a decimal number)
+        if answer.endswith('.') and not answer[-2].isdigit():
+            answer = answer[:-1].strip()
+        
+        # If answer starts with "The " and is followed by a name/noun, remove "The "
+        if answer.startswith("The ") and len(answer) > 4:
+            # Check if next word is capitalized (likely a proper noun)
+            next_word = answer.split()[1] if len(answer.split()) > 1 else ""
+            if next_word and next_word[0].isupper():
+                answer = answer[4:].strip()
+        
+        # Remove "a " or "an " from the beginning
+        if answer.lower().startswith("a "):
+            answer = answer[2:].strip()
+        elif answer.lower().startswith("an "):
+            answer = answer[3:].strip()
+        
+        print(f"[Cleaning] Original: '{original}' → Cleaned: '{answer}'")
+        return answer
+
+
+
+# Submission logic slightly changed
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """
-    Fetches all questions, runs the BasicAgent on them, submits all answers,
-    and displays the results.
+    Fetches all questions, runs the Premium Agent, and submits answers.
     """
     # --- Determine HF Space Runtime URL and Repo URL ---
     space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
@@ -67,15 +184,19 @@ 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)
+    # 1. Instantiate Agent (modified)
+    print("\n" + "="*30)
+    print("INITIALIZING PREMIUM AGENT")
+    print("="*30)
     try:
-        agent = BasicAgent()
+        agent = SmolGaiaAgent()
     except Exception as e:
         print(f"Error instantiating agent: {e}")
+        import traceback
+        traceback.print_exc()
         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}")
@@ -98,101 +219,117 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
         print(f"An unexpected error occurred fetching questions: {e}")
         return f"An unexpected error occurred fetching questions: {e}", None
 
-    # 3. Run your Agent
+    # 3. Run Agent with detailed progress tracking
     results_log = []
     answers_payload = []
-    print(f"Running agent on {len(questions_data)} questions...")
-    for item in questions_data:
+    total = len(questions_data)
+    
+    print("\n" + "="*30)
+    print(f"PROCESSING {total} QUESTIONS")
+    print("="*30 + "\n")
+    
+    for idx, item in enumerate(questions_data, 1):
         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}")
+            print(f"⚠ Skipping item with missing task_id or question")
             continue
+        
+        print(f"\n{'='*30}")
+        print(f"QUESTION {idx}/{total}")
+        print(f"Task ID: {task_id}")
+        print(f"Question: {question_text[:100]}...")
+        print('='*30)
+        
         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})
+            results_log.append({
+                "Task ID": task_id,
+                "Question": question_text,
+                "Submitted Answer": submitted_answer
+            })
+            print(f"✓ Answer recorded: {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}"})
+            print(f"✗ Error processing question: {e}")
+            import traceback
+            traceback.print_exc()
+            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)
 
-    # 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)
-
-    # 5. Submit
-    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
+        
+  
+
+    # 5. Submission
+    submission_data = {
+        "username": username.strip(),
+        "agent_code": agent_code,
+        "answers": answers_payload
+    }
+    
+    print("\n" + "="*30)
+    print(f"SUBMITTING {len(answers_payload)} ANSWERS")
+    print("="*30)
+    
     try:
         response = requests.post(submit_url, json=submission_data, timeout=60)
         response.raise_for_status()
         result_data = response.json()
+        
+        score = result_data.get('score', 'N/A')
+        correct = result_data.get('correct_count', '?')
+        total_attempted = result_data.get('total_attempted', '?')
+        
         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.')}"
+            f"Overall Score: {score}% ({correct}/{total_attempted} correct)\n"
+            f"Message: {result_data.get('message', 'No message received.')}\n\n"
+            f"{'EXCELLENT!' if float(score) >= 80 else 'Good job!' if float(score) >= 50 else 'Keep improving!'}"
         )
-        print("Submission successful.")
+        print(f"\n✓ Submission successful! Score: {score}%")
         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)
+        print(f"✗ Submission error: {e}")
         results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return f"Submission Failed: {e}", results_df
 
 
 # --- Build Gradio Interface using Blocks ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
+    gr.Markdown("# Premium Agent - Optimized for Maximum Accuracy")
     gr.Markdown(
         """
-        **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.  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.
+        **Current Configuration:**
+        - Model: Qwen/Qwen2.5-Coder-32B-Instruct (most capable)
+        - Max Steps: 12 (thorough reasoning)
+        - Enhanced answer cleaning
+        - Detailed progress logging
+        
+        **Target Performance:**
+        - Time: ~20-25 minutes for 20 questions
+        - Target Score: 60-80% (realistic for Level 1)
+        - Stretch Goal: 80%+ with optimal configuration
+        
+        **To Reach 100%:**
+        Getting 100% on GAIA Level 1 is extremely difficult. The benchmark shows:
+        - GPT-4 achieves ~70-80%
+        - Claude 3.5 achieves ~75-85%
+        - Human experts achieve ~90-95%
         """
     )
 
     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
+    run_button = gr.Button("Run Premium Evaluation & Submit")
+    status_output = gr.Textbox(label="Run Status / Submission Result", lines=7, interactive=False)
     results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
 
     run_button.click(
@@ -200,26 +337,19 @@ with gr.Blocks() as demo:
         outputs=[status_output, results_table]
     )
 
-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 Basic Agent Evaluation...")
+if __name__ == "__main__":
+    print("\n" + "="*30)
+    print("PREMIUM AGENT STARTING")
+    print("="*30)
+    
+    space_host = os.getenv("SPACE_HOST")
+    space_id = os.getenv("SPACE_ID")
+
+    if space_host:
+        print(f"✓ Runtime URL: https://{space_host}.hf.space")
+    if space_id:
+        print(f"✓ Repo URL: https://huggingface.co/spaces/{space_id}/tree/main")
+
+    print("="*30 + "\n")
     demo.launch(debug=True, share=False)