Update app.py

app.py

@@ -4,7 +4,6 @@ import requests
 import pandas as pd
 import asyncio
 import json
-import concurrent.futures
 from huggingface_hub import login
 from smolagents import CodeAgent, InferenceClientModel, DuckDuckGoSearchTool
 
@@ -19,8 +18,7 @@ login(token=os.environ["HUGGINGFACEHUB_API_TOKEN"])
 # --- Define Tools ---
 search_tool = DuckDuckGoSearchTool()
 
-
-# --- Main Function ---
+# --- Main Async Function ---
 async def run_and_submit_all(profile: gr.OAuthProfile | None):
     # Initialize Agent
     try:
@@ -31,9 +29,8 @@ async def run_and_submit_all(profile: gr.OAuthProfile | None):
             verbosity_level=2
         )
     except Exception as e:
-        return f"Error initializing agent: {e}", None
+        return f"❌ Agent Initialization Error: {e}", None
 
-    # Get Space ID for agent_code link
     space_id = os.getenv("SPACE_ID", "unknown")
     agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
 
@@ -41,18 +38,16 @@ async def run_and_submit_all(profile: gr.OAuthProfile | None):
     try:
         response = requests.get(QUESTIONS_URL, timeout=15)
         response.raise_for_status()
-        questions_data = response.json()
-        if not questions_data:
-            return "No questions received.", None
+        questions = response.json()
+        if not questions:
+            return "⚠️ No questions received.", None
     except Exception as e:
-        return f"Error fetching questions: {e}", None
+        return f"❌ Failed to fetch questions: {e}", None
 
-    # Prepare results
-    answers_payload = []
-    results_log = []
-    loop = asyncio.get_event_loop()
+    answers = []
+    logs = []
 
-    for item in questions_data:
+    for item in questions:
         task_id = item.get("task_id")
         question = item.get("question")
         if not task_id or not question:
@@ -61,100 +56,78 @@ async def run_and_submit_all(profile: gr.OAuthProfile | None):
         system_prompt = (
             "You are a general AI assistant. I will ask you a question. "
             "Report your thoughts, and finish your answer with the following template: "
-            "FINAL ANSWER: [YOUR FINAL ANSWER]. 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.\n\n"
+            "FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings.\n\n"
         )
-        prompt = system_prompt + f"Question: {question.strip()}"
+        full_prompt = system_prompt + f"Question: {question.strip()}"
 
-        # Run agent with timeout
         try:
-            with concurrent.futures.ThreadPoolExecutor() as executor:
-                future = executor.submit(agent, prompt)
-                agent_result = await loop.run_in_executor(None, future.result, 60)  # timeout=60s
-
-            # Clean model output
-            if isinstance(agent_result, dict) and "final_answer" in agent_result:
-                final_answer = str(agent_result["final_answer"]).strip()
-            elif isinstance(agent_result, str):
-                response_text = agent_result.strip()
-
-                # Remove known boilerplate
-                if "Here is the final answer from your managed agent" in response_text:
-                    response_text = response_text.split(":", 1)[-1].strip()
-
-                # Extract final answer
-                if "FINAL ANSWER:" in response_text:
-                    _, final_answer = response_text.rsplit("FINAL ANSWER:", 1)
-                    final_answer = final_answer.strip()
+            loop = asyncio.get_running_loop()
+            result = await loop.run_in_executor(None, lambda: agent(full_prompt))
+
+            if isinstance(result, dict) and "final_answer" in result:
+                final_answer = str(result["final_answer"]).strip()
+            elif isinstance(result, str):
+                if "FINAL ANSWER:" in result:
+                    final_answer = result.split("FINAL ANSWER:")[-1].strip()
                 else:
-                    final_answer = response_text
+                    final_answer = result.strip()
             else:
-                final_answer = str(agent_result).strip()
+                final_answer = str(result).strip()
 
         except Exception as e:
             print(f"[ERROR] Task {task_id} failed: {e}")
             final_answer = f"AGENT ERROR: {e}"
 
-        answers_payload.append({"task_id": task_id, "model_answer": final_answer})
-        results_log.append({"Task ID": task_id, "Question": question, "Submitted Answer": final_answer})
+        answers.append({"task_id": task_id, "model_answer": final_answer})
+        logs.append({"Task ID": task_id, "Question": question, "Submitted Answer": final_answer})
 
-    # Clean invalid entries
-    valid_answers = [a for a in answers_payload if isinstance(a["task_id"], str) and isinstance(a["model_answer"], str)]
+    valid_answers = [a for a in answers if isinstance(a["task_id"], str) and isinstance(a["model_answer"], str)]
 
     if not valid_answers:
-        return "Agent produced no valid answers.", pd.DataFrame(results_log)
+        return "❌ Agent produced no valid answers.", pd.DataFrame(logs)
 
-    # Prepare submission
-    username = profile.username if profile else "unknown"
-    submission_data = {
-        "username": username.strip(),
+    submission = {
+        "username": profile.username if profile else "unknown",
         "agent_code": agent_code,
         "answers": valid_answers
     }
 
-    print("[DEBUG] Submission Payload:\n", json.dumps(submission_data, indent=2))
+    print("[DEBUG] Submitting:\n", json.dumps(submission, indent=2))
 
     try:
-        response = requests.post(SUBMIT_URL, json=submission_data, timeout=60)
-        response.raise_for_status()
-        result_data = response.json()
+        resp = requests.post(SUBMIT_URL, json=submission, timeout=60)
+        resp.raise_for_status()
+        result_data = resp.json()
 
-        final_status = (
+        summary = (
             f"✅ Submission Successful\n"
             f"User: {result_data.get('username')}\n"
             f"Score: {result_data.get('score', 'N/A')}% "
             f"({result_data.get('correct_count')}/{result_data.get('total_attempted')})\n"
             f"Message: {result_data.get('message', 'No message.')}"
         )
-        return final_status, pd.DataFrame(results_log)
+        return summary, pd.DataFrame(logs)
 
     except Exception as e:
-        return f"Submission Failed: {e}", pd.DataFrame(results_log)
-
+        return f"❌ Submission failed: {e}", pd.DataFrame(logs)
 
 # --- Gradio UI ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Agent Evaluation Interface")
+    gr.Markdown("# 🧠 GAIA Agent Evaluation Interface")
     gr.Markdown("""
-    **Instructions:**
-    1. Clone and customize the agent logic.
-    2. Log in to Hugging Face.
-    3. Click "Run Evaluation" to test and submit your answers.
+    - Log in with your Hugging Face account.
+    - Click the button below to run the agent and submit the answers.
+    - Wait for the final score to appear.
     """)
 
     gr.LoginButton()
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
-    status_output = gr.Textbox(label="Status", lines=5, interactive=False)
-    results_table = gr.DataFrame(label="Agent Answers", wrap=True)
+    run_button = gr.Button("🚀 Run Evaluation & Submit")
+    status = gr.Textbox(label="Status", lines=6)
+    table = gr.DataFrame(label="Answer Log")
 
-    run_button.click(fn=run_and_submit_all, outputs=[status_output, results_table])
-#hi
+    run_button.click(fn=run_and_submit_all, outputs=[status, table])
 
-# --- App Launch ---
+# --- Launch ---
 if __name__ == "__main__":
-    print("\n--- Launching Gradio Space ---")
-    print(f"✅ SPACE_HOST: {os.getenv('SPACE_HOST')}")
-    print(f"✅ SPACE_ID: {os.getenv('SPACE_ID')}")
-    demo.launch(debug=True, share=False)
+    print("Launching Agent Space...")
+    demo.launch(debug=True)