Update app.py

app.py

@@ -11,14 +11,9 @@ from torch import cuda
 import json
 from datetime import datetime
 
-
+model_config = getconfig("model_params.cfg")
 ###########
-# ToDo move to config file and functions
-# Configuration for the dedicated model
-DEDICATED_MODEL = "meta-llama/Llama-3.1-8B-Instruct"
-DEDICATED_ENDPOINT = "https://qu2d8m6dmsollhly.us-east-1.aws.endpoints.huggingface.cloud"
-# Retrieve your write access token from the settings (assuming you stored it in st.secrets)
-WRITE_ACCESS_TOKEN = st.secrets["Llama_3_1"]
+# ToDo move to functions
 
 def get_rag_answer(query, top_results):
     """
@@ -260,8 +255,12 @@ if show_exact_matches:
     if not filtered_lexical_no_dupe:
         st.write('No exact matches, consider unchecking "Show only exact matches"')
     else:
-        # 6) Display the first 15 matching results
-        for res in filtered_lexical_no_dupe[:15]:
+        top_results = filtered_lexical_no_dupe[:15]
+        rag_answer = get_rag_answer(var, top_results)
+        st.markdown("### Generated Answer")
+        st.write(rag_answer)
+        st.divider()
+        for res in top_results:
             # Metadata
             metadata = res.payload.get('metadata', {})
             countries = metadata.get('countries', "[]")