update

demo.py

@@ -213,7 +213,7 @@ def ensure_client_available():
 def model_prompting(
     llm_model: str,
     prompt: str,
-    max_token_num: Optional[int] = 2048,
+    max_token_num: Optional[int] = 1024,  # Changed from 2048 to 1024
     temperature: Optional[float] = 0.2,
     top_p: Optional[float] = 0.7,
     stream: Optional[bool] = True,
@@ -766,7 +766,7 @@ class GNNLLMSystem:
             response = model_prompting(
                 llm_model=mapped_model_name,
                 prompt=formatted_query,
-                max_token_num=4096,
+                max_token_num=1024,  # Changed from 4096 to 1024
                 temperature=0.0,
                 top_p=0.9,
                 stream=True
@@ -859,7 +859,7 @@ Keep the description concise and informative. Respond with just the task descrip
         task_description = model_prompting(
             llm_model="meta/llama-3.1-8b-instruct",
             prompt=prompt,
-            max_token_num=2048,
+            max_token_num=1024,  # Changed from 2048 to 1024
             temperature=0.1,
             top_p=0.9,
             stream=True
@@ -1073,7 +1073,7 @@ def process_query(query):
         response = model_prompting(
             llm_model=api_model,
             prompt=final_prompt,
-            max_token_num=4096,
+            max_token_num=1024,  # Changed from 4096 to 1024
             temperature=0.0,
             top_p=0.9,
             stream=True
@@ -1187,7 +1187,7 @@ def process_thought_template_query(query, template_style, task_description, top_
             llama_response = model_prompting(
                 llm_model="meta/llama-3.1-8b-instruct",
                 prompt=enhanced_query,
-                max_token_num=4096,
+                max_token_num=1024,  # Changed from 4096 to 1024
                 temperature=0.0,
                 top_p=0.9,
                 stream=True