fix: couple of issues

app.py

@@ -31,11 +31,10 @@ DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 # ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
 class BasicAgent:
     def __init__(self):
+        # TODO inspect messages exchanged between the llm and the agent
+
         # self.llm = Ollama(model="qwen2.5:7b", request_timeout=500)
         self.llm = Gemini(model_name="models/gemini-2.0-flash")
-        print("BasicAgent initialized.")
-    def __call__(self, question: str) -> str:
-        print(f"Agent received question (first 50 chars): {question[:50]}...")
         def get_answers_from_youtube_transcript(youtube_link: str) -> str:
             """Fetches transcript of the given youtube_link and returns matching answers based on query.
                 Args:
@@ -70,7 +69,7 @@ class BasicAgent:
                 nodes = splitter.get_nodes_from_documents(documents)
 
                 retriever = BM25Retriever(nodes=nodes, similarity_top_k=10)
-                synthesizer = get_response_synthesizer(response_mode="refine", llm=self.llm)
+                synthesizer = get_response_synthesizer(response_mode="tree_summarize", llm=self.llm)
                 query_engine = RetrieverQueryEngine(retriever=retriever, response_synthesizer=synthesizer)
 
                 response = query_engine.query(query)
@@ -95,7 +94,7 @@ class BasicAgent:
                 nodes = splitter.get_nodes_from_documents(documents)
 
                 retriever = BM25Retriever(nodes=nodes, similarity_top_k=1)
-                synthesizer = get_response_synthesizer(response_mode="refine", llm=llm)
+                synthesizer = get_response_synthesizer(response_mode="tree_summarize", llm=self.llm)
                 query_engine = RetrieverQueryEngine(retriever=retriever, response_synthesizer=synthesizer)
 
                 response = query_engine.query(query)
@@ -111,17 +110,22 @@ class BasicAgent:
                 description="Searches wikipedia and converts results into a high quality answer."
             )
 
-        agent = AgentWorkflow.from_tools_or_functions([duckduckgo_search_tool, youtube_transcript_answer_tool, wikipedia_search_tool], llm=self.llm)
-        async def run_agent():
-            return await agent.run(question)
+        self.agent = AgentWorkflow.from_tools_or_functions([duckduckgo_search_tool, youtube_transcript_answer_tool, wikipedia_search_tool], llm=self.llm)
+        print("BasicAgent initialized.")
+
+    async def run_agent(self, question: str):
+        return await self.agent.run(question)
+    
+    def __call__(self, question: str) -> str:
+        print(f"Agent received question (first 50 chars): {question[:50]}...")
         
-        response = asyncio.run(run_agent())
+        response = asyncio.run(self.run_agent(question=question))
 
         final_answer = response.response.blocks[0].text
         print(f"Agent returning fixed answer: {final_answer}")
         return final_answer
 
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+async def run_and_submit_all( profile: gr.OAuthProfile | None):
     """
     Fetches all questions, runs the BasicAgent on them, submits all answers,
     and displays the results.
@@ -176,7 +180,7 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
     answers_payload = []
     print(f"Running agent on {len(questions_data)} questions...")
     for item in questions_data:
-        time.sleep(60)
+        await asyncio.sleep(60)
         task_id = item.get("task_id")
         question_text = item.get("question")
         if not task_id or question_text is None: