bad param

app.py

@@ -39,8 +39,8 @@ class SlpMultiAgent:
         model = OpenAIServerModel(
             model_id="gpt-3.5-turbo-16k", 
             temperature=0.0, 
-            max_tokens=1000,
-            request_timeout=60
+            max_tokens=1000
+            # Removed request_timeout parameter
         )
         
         # Here you can implement your agent logic, tools, and model calls
@@ -58,8 +58,8 @@ class SlpMultiAgent:
             model=OpenAIServerModel(
                 model_id="gpt-3.5-turbo-16k",
                 temperature=0.0,
-                max_tokens=1000,
-                request_timeout=60
+                max_tokens=1000
+                # Removed request_timeout parameter
             ),
             tools=[],
             managed_agents=[web_agent],
@@ -125,8 +125,8 @@ def check_reasoning(final_answer, agent_memory):
     try:
         multimodal_model = OpenAIServerModel(
             model_id="gpt-3.5-turbo", 
-            max_tokens=500,
-            request_timeout=30
+            max_tokens=500
+            # Removed request_timeout parameter
         )
         
         # Simplified prompt to reduce token usage
@@ -221,7 +221,8 @@ async def run_and_submit_all(profile):
     answers_payload = []
     print(f"Running agent on {len(questions_data)} questions...")
     
-    semaphore = asyncio.Semaphore(3)
+    # Process questions with lower concurrency to avoid rate limits
+    semaphore = asyncio.Semaphore(1)  # Process one question at a time
     
     async def process_question(item):
         task_id = item.get("task_id")