Try to switch to llama model to see if it handles more tokens

app.py

@@ -26,7 +26,8 @@ if st.button("Submit"):
         ]
 
         completion = client.chat.completions.create(
-            model="Qwen/Qwen2.5-Coder-32B-Instruct", 
+            #model="Qwen/Qwen2.5-Coder-32B-Instruct", 
+            model="meta-llama/LLaMA-3-8b",
         	messages=messages, 
         	max_tokens=500
         )