Update app.py

app.py

@@ -1,211 +1,136 @@
 
+app/py
 import os
+import gradio as gr
 import requests
 import pandas as pd
-import gradio as gr
+from smolagents import CodeAgent, DuckDuckGoSearchTool
+from smolagents.models import OpenAIServerModel
+import openai
 
-from smolagents import ToolCallingAgent, OpenAIServerModel
-from audio_transcriber import AudioTranscriptionTool
-from image_analyzer import ImageAnalysisTool
-from wikipedia_searcher import WikipediaSearcher
+# --- Setup ---
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+if not OPENAI_API_KEY:
+    raise RuntimeError("Please set OPENAI_API_KEY in your Space secrets.")
+openai.api_key = OPENAI_API_KEY
 
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+OPENAI_MODEL_ID = os.getenv("OPENAI_MODEL_ID", "gpt-4o")
 
-class GaiaAgent:
-    def __init__(self):
-        tools = [
-            AudioTranscriptionTool(),
-            ImageAnalysisTool(),
-            WikipediaSearcher()
-        ]
-
-        model_id = os.getenv("OPENAI_MODEL_ID", "gpt-3.5-turbo")
-        self.agent = ToolCallingAgent(
-            model=OpenAIServerModel(model_id=model_id),
-            tools=tools
-        )
+model = OpenAIServerModel(model_id=OPENAI_MODEL_ID, api_key=OPENAI_API_KEY)
+search_tool = DuckDuckGoSearchTool()
+agent = CodeAgent(tools=[search_tool], model=model)
 
-    def __call__(self, question: str) -> str:
-        prompt = f"""You are an agent solving the GAIA benchmark and you are required to provide exact answers.
-        Rules to follow:
-        1. Return only the exact requested answer: no explanation and no reasoning.
-        2. For yes/no questions, return exactly "Yes" or "No".
-        3. For dates, use the exact format requested.
-        4. For numbers, use the exact number, no other format.
-        5. For names, use the exact name as found in sources.
-        6. If the question has an associated file, download the file first using the task ID.
-        Examples of good responses:
-        - "42"
-        - "Pinco Palla"
-        - "Yes"
-        - "October 5, 2001"
-        - "Buenos Aires"
-        Never include phrases like "the answer is..." or "Based on my research".
-        Only return the exact answer.
-        QUESTION:
-        {question}
-        """
-        result = self.agent.run(prompt)
-        return result  # ✅ Fixed: removed .get() since result is a string
-
-
-def run_and_submit_all(profile: gr.OAuthProfile | None):
-    space_id = os.getenv("SPACE_ID")
-
-    if profile:
-        username = profile.username
-        if isinstance(username, list):
-            username = username[0]
-        username = username.strip()
-        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"
+answer_formatting_prompt = """
+You are a smart assistant with access to tools like DuckDuckGoSearchTool(query: str). 
+Think step-by-step, then output your response.
 
-    try:
-        agent = GaiaAgent()
-    except Exception as e:
-        print(f"Error initializing agent: {e}")
-        return f"Error initializing agent: {e}", None
+IMPORTANT: 
+FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers/strings. 
+Do NOT include commas, $ or % unless asked.
+Write digits plainly (e.g., '10', not 'ten').
+
+Use format:
+FINAL ANSWER: <your_answer>
+"""
+
+def show_profile(profile):
+    if not profile:
+        return "⚠️ Not logged in."
+    return f"✅ Logged in as: {profile['username']}"
 
+def run_and_submit_all(login_info):
+    # login_info comes from LoginButton, it's None if not logged in
+    if not login_info:
+        return "⚠️ Please log in with your Hugging Face account.", pd.DataFrame()
+
+    username = login_info["username"]
+    space_id = os.getenv("SPACE_ID", "")
     agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(f"Agent code URL: {agent_code}")
 
     try:
-        response = requests.get(questions_url, timeout=15)
-        response.raise_for_status()
-        questions_data = response.json()
-        if not questions_data:
-            return "Fetched questions list is empty or invalid format.", None
-        print(f"Fetched {len(questions_data)} questions.")
+        resp = requests.get(f"{DEFAULT_API_URL}/questions", timeout=15)
+        resp.raise_for_status()
+        questions = resp.json()
     except Exception as e:
-        return f"Error fetching questions: {e}", None
-
-    results_log = []
-    answers_payload = []
+        return f"❌ Error fetching questions: {e}", pd.DataFrame()
 
-    for item in questions_data:
+    results, payload = [], []
+    for item in questions:
         task_id = item.get("task_id")
-        if not task_id:
+        question = item.get("question")
+        if not task_id or not question:
             continue
-
-        question_text = item.get("question", "")
-
-        file_url = item.get("file_url")
-        local_file_path = None
-        if file_url:
-            try:
-                ext = file_url.split(".")[-1].lower()
-                if ext in ["mp3", "wav", "jpeg", "jpg", "png"]:
-                    local_file_path = f"./temp_{task_id}.{ext}"
-                    with requests.get(file_url, stream=True) as r:
-                        r.raise_for_status()
-                        with open(local_file_path, "wb") as f:
-                            for chunk in r.iter_content(chunk_size=8192):
-                                f.write(chunk)
-                    print(f"Downloaded file for task {task_id} to {local_file_path}")
-                    question_text += f"\n\nFile path: {local_file_path}"
-            except Exception as e:
-                print(f"Failed to download file for task {task_id}: {e}")
-
+        prompt = answer_formatting_prompt.strip() + f"\n\nQUESTION: {question.strip()}"
         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
-            })
+            answer = agent.run(prompt)
         except Exception as e:
-            error_msg = f"AGENT ERROR: {e}"
-            results_log.append({
-                "Task ID": task_id,
-                "Question": question_text,
-                "Submitted Answer": error_msg
-            })
-
-        if local_file_path:
-            try:
-                os.remove(local_file_path)
-            except Exception:
-                pass
-
-    if not answers_payload:
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
-
-    submission_data = {
-        "username": username,
-        "agent_code": agent_code,
-        "answers": answers_payload
-    }
-
-    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
+            answer = f"AGENT ERROR: {e}"
+        results.append({"Task ID": task_id, "Question": question, "Submitted Answer": answer})
+        payload.append({"task_id": task_id, "submitted_answer": answer})
+
+    if not payload:
+        return "⚠️ Agent returned no answers.", pd.DataFrame(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.')}"
+        post = requests.post(
+            f"{DEFAULT_API_URL}/submit",
+            json={"username": username, "agent_code": agent_code, "answers": payload},
+            timeout=60
+        )
+        post.raise_for_status()
+        result = post.json()
+        score = result.get("score", "N/A")
+        correct = result.get("correct_count", "?")
+        attempted = result.get("total_attempted", "?")
+        message = result.get("message", "")
+        return (
+            f"✅ Submission Successful!\nUser: {username}\nScore: {score}% "
+            f"({correct}/{attempted})\nMessage: {message}",
+            pd.DataFrame(results)
         )
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
-    except requests.exceptions.HTTPError as e:
-        try:
-            detail = e.response.json().get("detail", e.response.text)
-        except Exception:
-            detail = e.response.text[:500]
-        return f"Submission Failed: {detail}", pd.DataFrame(results_log)
-    except requests.exceptions.Timeout:
-        return "Submission Failed: The request timed out.", pd.DataFrame(results_log)
     except Exception as e:
-        return f"An unexpected error occurred during submission: {e}", pd.DataFrame(results_log)
+        return f"❌ Submission failed: {e}", pd.DataFrame(results)
 
 
-# Gradio UI
 with gr.Blocks() as demo:
     gr.Markdown("# Basic Agent Evaluation Runner")
-    gr.Markdown("""\
-**Instructions:**
-1. Clone this space and define your agent and tools.
-2. Log in to your Hugging Face account using the button below.
-3. Click 'Run Evaluation & Submit All Answers' to test your agent and submit results.
-""")
-
-    gr.LoginButton()
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
 
+    login_button = gr.LoginButton()
+    login_status = gr.Textbox(label="Login Status")
+
+    run_button = gr.Button("Run Evaluation & Submit All Answers")
     status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
     results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
 
-    run_button.click(fn=run_and_submit_all, outputs=[status_output, results_table])
+    # Show login status when user logs in
+    login_button.click(fn=show_profile, inputs=[login_button], outputs=[login_status])
+
+    # Run evaluation on click, pass login_button's state as input
+    run_button.click(fn=run_and_submit_all, inputs=[login_button], outputs=[status_output, results_table])
+
 
 if __name__ == "__main__":
-    print("\n" + "-"*30 + " App Starting " + "-"*30)
-    space_host = os.getenv("SPACE_HOST")
-    space_id = os.getenv("SPACE_ID")
-
-    if space_host:
-        print(f"✅ SPACE_HOST found: {space_host}")
-        print(f"   Runtime URL should be: https://{space_host}.hf.space")
-    else:
-        print("ℹ️  SPACE_HOST not found.")
-
-    if space_id:
-        print(f"✅ SPACE_ID found: {space_id}")
-        print(f"   Repo URL: https://huggingface.co/spaces/{space_id}")
-    else:
-        print("ℹ️  SPACE_ID not found.")
-
-    print("-"*(60 + len(" App Starting ")) + "\n")
-    demo.launch(debug=True, share=False)
+    demo.launch()
+
+
+#import gradio as gr
+
+#def show_profile(profile):
+ #   if not profile:
+   #     return "⚠️ Not logged in."
+  #  return f"✅ Logged in as: {profile['username']}"
+
+ # with gr.Blocks() as demo:
+   #   gr.Markdown("## 🔐 Hugging Face OAuth Login")
+
+   #   login_button = gr.LoginButton()
+   #   output = gr.Textbox(label="Login Status")
+
+    #  login_button.click(fn=show_profile, inputs=[login_button], outputs=[output])
+
+ # demo.launch()
+