some tweaks

app.py

@@ -2,7 +2,7 @@ import os
 from langchain_huggingface import HuggingFaceEndpoint
 import streamlit as st
 model_id="mistralai/Mistral-7B-Instruct-v0.3"
-def get_llm_hf_inference(model_id=model_id, max_new_tokens=128, temperature=0.1):
+def get_llm_hf_inference(model_id=model_id, max_new_tokens=4096, temperature=0.1):
     """
     Returns a language model for HuggingFace inference.
 
@@ -40,7 +40,7 @@ if 'user_text' not in st.session_state:
 
 # Initialize session state for model parameters
 if "max_response_length" not in st.session_state:
-    st.session_state.max_response_length = 128
+    st.session_state.max_response_length = 4096
 
 if "system_message" not in st.session_state:
     st.session_state.system_message = "rude AI conversing with a human user"