Update app.py

app.py

@@ -5,154 +5,146 @@ import re
 import pandas as pd
 from transformers import pipeline, set_seed
 
-# --- Constants ---
+# ---------------- CONSTANTS ----------------
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- Smarter Basic Agent ---
+# ---------------- BASIC AGENT ----------------
 class BasicAgent:
     def __init__(self):
-        print("BasicAgent initialized.")
+        print("BasicAgent initialized")
         set_seed(42)
+
         self.generator = pipeline(
             "text2text-generation",
             model="google/flan-t5-large",
-            max_new_tokens=32
+            max_new_tokens=32,
+            temperature=0.0,      # VERY IMPORTANT: deterministic
+            do_sample=False
         )
 
     def __call__(self, question: str) -> str:
-        # --- Detect question type ---
-        q_lower = question.lower()
-
-        # Boolean detection
-        if any(x in q_lower for x in ["true or false", "yes or no", "is it", "does it", "can it"]):
-            prompt = (
-                "Answer only True or False exactly. "
-                "No explanations or extra words.\n"
-                f"Q: {question}\nA:"
-            )
-        # Multiple choice detection
-        elif any(x in question for x in ["A)", "B)", "C)", "D)", "choose", "options"]):
-            prompt = (
-                "Answer only with one option exactly: A, B, C, or D. "
-                "No explanations or extra words.\n"
-                f"Q: {question}\nA:"
-            )
-        # Numeric detection (digits in question or ask for a number)
-        elif re.search(r"number|digits|how many|count|what is \d", q_lower):
-            prompt = (
-                "Answer only with the number, no words or punctuation.\n"
-                f"Q: {question}\nA:"
-            )
-        # Default: let model generate
-        else:
-            prompt = (
-                "Give ONLY the exact answer. "
-                "No explanation or extra words.\n"
-                f"Q: {question}\nA:"
-            )
-
-        # --- Generate answer ---
+        prompt = (
+            "Answer the question with ONLY the final answer.\n"
+            "No explanation. No extra words. No punctuation.\n\n"
+            f"Question: {question}\n"
+            "Answer:"
+        )
+
         result = self.generator(prompt)[0]["generated_text"]
 
-        # --- Post-process answer ---
+        # ---------- STRONG CLEANUP ----------
         answer = result.strip()
-        answer = re.sub(r"(?i)^(the answer is|answer:)\s*", "", answer)  # remove prefixes
-        answer = re.split(r"[.\n]", answer)[0].strip()  # first line or sentence
-        answer = answer.rstrip(".,;:")  # remove trailing punctuation
-
-        # Enforce boolean formatting
-        if answer.lower() in ["true", "false"]:
-            answer = answer.capitalize()
-
-        print(f"Question: {question}\nFinal answer: {answer}\n")
+        answer = re.sub(r"(?i)^(the answer is|answer:)\s*", "", answer)
+        answer = re.split(r"[\n\.]", answer)[0]
+        answer = answer.strip()
+        answer = answer.rstrip(".,;:")
+
+        # Normalize boolean answers
+        if answer.lower() == "true":
+            answer = "True"
+        elif answer.lower() == "false":
+            answer = "False"
+
+        print(f"\nQ: {question}\nA: {answer}\n")
         return answer
 
 
-# --- Run and Submit Function ---
+# ---------------- RUN + SUBMIT ----------------
 def run_and_submit_all(profile: gr.OAuthProfile | None):
     space_id = os.getenv("SPACE_ID")
 
-    if profile:
-        username = profile.username
-    else:
-        return "Please Login to Hugging Face with the button.", pd.DataFrame()
+    if not profile:
+        return "Please login with Hugging Face.", pd.DataFrame()
 
-    api_url = DEFAULT_API_URL
-    questions_url = f"{api_url}/questions"
-    submit_url = f"{api_url}/submit"
+    username = profile.username
+    agent = BasicAgent()
 
-    try:
-        agent = BasicAgent()
-    except Exception as e:
-        return f"Error initializing agent: {e}", pd.DataFrame()
+    agent_code = (
+        f"https://huggingface.co/spaces/{space_id}/tree/main"
+        if space_id else "N/A"
+    )
 
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main" if space_id else "Agent code URL not available."
-
-    # --- Fetch Questions ---
+    # Fetch questions
     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.", pd.DataFrame()
+        questions = requests.get(
+            f"{DEFAULT_API_URL}/questions", timeout=15
+        ).json()
     except Exception as e:
-        return f"Error fetching questions: {e}", pd.DataFrame()
+        return f"Failed to fetch questions: {e}", pd.DataFrame()
 
-    results_log = []
     answers_payload = []
+    results_log = []
+
+    for q in questions:
+        task_id = q["task_id"]
+        question_text = q["question"]
 
-    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:
-            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
-            })
+            answer = agent(question_text)
         except Exception as e:
-            results_log.append({
-                "Task ID": task_id,
-                "Question": question_text,
-                "Submitted Answer": f"AGENT ERROR: {e}"
-            })
-
-    if not answers_payload:
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
-
-    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
-
+            answer = "ERROR"
+
+        answers_payload.append({
+            "task_id": task_id,
+            "submitted_answer": answer
+        })
+
+        results_log.append({
+            "Task ID": task_id,
+            "Question": question_text,
+            "Submitted Answer": answer
+        })
+
+    submission_data = {
+        "username": username,
+        "agent_code": agent_code,
+        "answers": answers_payload
+    }
+
+    # Submit answers
     try:
-        response = requests.post(submit_url, json=submission_data, timeout=60)
-        response.raise_for_status()
-        result_data = response.json()
-        final_status = (
+        response = requests.post(
+            f"{DEFAULT_API_URL}/submit",
+            json=submission_data,
+            timeout=60
+        ).json()
+
+        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"User: {response.get('username')}\n"
+            f"Overall Score: {response.get('score')}% "
+            f"({response.get('correct_count')}/"
+            f"{response.get('total_attempted')} correct)\n"
+            f"Message: {response.get('message')}"
         )
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
     except Exception as e:
-        return f"Submission Failed: {e}", pd.DataFrame(results_log)
+        status = f"Submission failed: {e}"
+
+    return status, pd.DataFrame(results_log)
 
 
-# --- Gradio Interface ---
+# ---------------- GRADIO UI ----------------
 with gr.Blocks() as demo:
-    gr.Markdown("# GAIA Smarter Agent Runner")
+    gr.Markdown("# GAIA Unit 4 – Basic Agent Runner")
+
     gr.LoginButton()
-    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_btn = gr.Button("Run Evaluation & Submit")
+
+    status_box = gr.Textbox(
+        label="Submission Result",
+        lines=5,
+        interactive=False
+    )
 
-    run_button.click(fn=run_and_submit_all, outputs=[status_output, results_table])
+    results_table = gr.DataFrame(
+        label="Questions and Answers",
+        wrap=True
+    )
 
+    run_btn.click(
+        fn=run_and_submit_all,
+        outputs=[status_box, results_table]
+    )
 
 if __name__ == "__main__":
-    demo.launch(debug=True, share=False)
+    demo.launch(debug=True)