Update app.py

app.py

@@ -1,145 +1,102 @@
-# app.py  –  vollständige, lauffähige Fassung
-# -------------------------------------------
-import os
-import gradio as gr
-import requests
-import pandas as pd
+# app.py  –  async + progress, keine Cache-Logik
+# ------------------------------------------------
+import os, asyncio, concurrent.futures, functools
+import gradio as gr, requests, pandas as pd
+from langchain_core.messages import HumanMessage
+from agent import agent_executor     # dein LangGraph-Agent
 
-from agent import agent_executor                  # dein LangGraph-Agent
-from langchain_core.messages import HumanMessage  # NEU: benötigt für llm_input
 
-
-# (Keep Constants as is)
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+MAX_PAR_TASKS   = 5                  # wie viele Fragen parallel laufen
 
 
-# ---------------------------------------------------------------------------
-#  BasicAgent-Wrapper: ruft den LangGraph-Executor auf
-# ---------------------------------------------------------------------------
-class BasicAgent:
-    def __init__(self):
-        print("LLM Tool-Enhanced Agent initialized.")
-
-    # nimmt jetzt ein Dict (messages + task_id) entgegen
-    def __call__(self, llm_input: dict) -> str:
-        try:
-            result = agent_executor.invoke(llm_input)   # LangGraph ausführen
-            answer = result["messages"][-1].content
-            return answer.strip()
-        except Exception as e:
-            print(f"Agent error: {e}")
-            return "I don't know."
-
-
-# ---------------------------------------------------------------------------
-#  GAIA-Runner: Fragen holen → Agent laufen lassen → Ergebnis submitten
-# ---------------------------------------------------------------------------
-def run_and_submit_all(profile: gr.OAuthProfile | None):
-    """Fetch GAIA questions, run agent, submit answers."""
-    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"
-
-    # Agent instanziieren
+# ------------------------------------------------------------------
+# Sync-Wrapper um den Agent
+# ------------------------------------------------------------------
+def run_agent_sync(task_id: str, question: str) -> str:
+    llm_input = {
+        "messages": [HumanMessage(content=question)],
+        "task_id":  task_id,
+    }
     try:
-        agent = BasicAgent()
+        result = agent_executor.invoke(llm_input)
+        return result["messages"][-1].content.strip()
     except Exception as e:
-        return f"Error initializing agent: {e}", None
+        return f"AGENT ERROR: {e}"
+
+
+async def run_agent_async(executor, task_id: str, question: str) -> str:
+    loop = asyncio.get_event_loop()
+    return await loop.run_in_executor(
+        executor, functools.partial(run_agent_sync, task_id, question)
+    )
 
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(agent_code)
 
-    # Fragen holen
+# ------------------------------------------------------------------
+# Haupt-Callback (async)  –  holt Fragen, verarbeitet parallel
+# ------------------------------------------------------------------
+async def run_and_submit_all(profile: gr.OAuthProfile | None, progress=gr.Progress()):
+    if not profile:
+        return "Please login with your HF account.", None
+    username = profile.username
+
+    # 1) GAIA-Fragen holen
+    q_url = f"{DEFAULT_API_URL}/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.", None
-        print(f"Fetched {len(questions_data)} questions.")
+        q_data = requests.get(q_url, timeout=15).json()
     except Exception as e:
         return f"Error fetching questions: {e}", None
 
-    # Agent auf jede Frage anwenden
-    results_log, answers_payload = [], []
-    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:
-            llm_input = {
-                "messages": [HumanMessage(content=question_text)],
-                "task_id": task_id,              # ←  WICHTIG!
-            }
-            submitted_answer = agent(llm_input)
-
-            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:
-            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
-    submission_data = {
-        "username": username.strip(),
-        "agent_code": agent_code,
-        "answers": answers_payload,
+    progress(0, desc=f"Fetched {len(q_data)} questions – processing …")
+
+    # 2) Parallel ausführen
+    answers, log_rows = [], []
+    with concurrent.futures.ThreadPoolExecutor(max_workers=MAX_PAR_TASKS) as ex:
+        tasks = [
+            run_agent_async(ex, itm["task_id"], itm["question"])
+            for itm in q_data
+        ]
+        for i, coro in enumerate(asyncio.as_completed(tasks), 1):
+            answer = await coro
+            task_id = q_data[i-1]["task_id"]
+            question = q_data[i-1]["question"]
+
+            answers.append({"task_id": task_id, "submitted_answer": answer})
+            log_rows.append({"Task ID": task_id, "Question": question, "Answer": answer})
+
+            progress(i / len(q_data), desc=f"{i}/{len(q_data)} done")
+
+    # 3) Antworten submitten
+    submit_url = f"{DEFAULT_API_URL}/submit"
+    payload = {
+        "username": username,
+        "agent_code": f"https://huggingface.co/spaces/{os.getenv('SPACE_ID')}/tree/main",
+        "answers": answers,
     }
     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', '?')}/"
-            f"{result_data.get('total_attempted', '?')} correct)\n"
-            f"Message: {result_data.get('message', 'No message received.')}"
+        res = requests.post(submit_url, json=payload, timeout=60).json()
+        status = (
+            f"Submission OK – Score: {res.get('score','?')} % "
+            f"({res.get('correct_count','?')}/{res.get('total_attempted','?')})"
         )
-        return final_status, pd.DataFrame(results_log)
-
     except Exception as e:
-        status_message = f"Submission Failed: {e}"
-        return status_message, pd.DataFrame(results_log)
+        status = f"Submission failed: {e}"
 
+    return status, pd.DataFrame(log_rows)
 
-# ---------------------------------------------------------------------------
-#  Gradio-UI (unverändert)
-# ---------------------------------------------------------------------------
-with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
 
+# ------------------------------------------------------------------
+# Gradio-UI
+# ------------------------------------------------------------------
+with gr.Blocks() as demo:
+    gr.Markdown("# Fast GAIA Agent Runner  (Async + Progress)")
     gr.LoginButton()
-    run_button = gr.Button("Run Evaluation & Submit All Answers")
+    run_btn = gr.Button("Run & Submit")
 
-    status_output = gr.Textbox(label="Run Status / Submission Result",
-                               lines=5, interactive=False)
-    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
+    out_status = gr.Textbox(label="Status / Score", lines=3, interactive=False)
+    out_table  = gr.DataFrame(label="Answers", wrap=True)
 
-    run_button.click(fn=run_and_submit_all,
-                     outputs=[status_output, results_table])
+    run_btn.click(run_and_submit_all, outputs=[out_status, out_table])
 
 if __name__ == "__main__":
     demo.launch(debug=True, share=False)