Update app.py

app.py

@@ -25,7 +25,7 @@ import os
 from io import StringIO
 from llama_index.core.memory import ChatMemoryBuffer
 
-memory = ChatMemoryBuffer.from_defaults(token_limit=3072)
+memory = ChatMemoryBuffer.from_defaults(token_limit=1500)
 #
 
 last = 0
@@ -82,7 +82,7 @@ client = InferenceClient(LLM_NAME)
 Settings.llm = HuggingFaceInferenceAPI(model_name=LLM_NAME,
 #                              device_map="auto",
 #                              system_prompt = system_prompt,
-                              context_window=6144,
+                              context_window=4092,
                               max_new_tokens=3072,
 #                              stopping_ids=[50278, 50279, 50277, 1, 0],
                               generate_kwargs={"temperature": temperature, "top_p":top_p, "repetition_penalty": repetition_penalty,