Update app.py

app.py

@@ -160,11 +160,9 @@ def generate_response(query,  model, template):
     progress_text = "Running Inference. Please wait."
     my_bar = st.progress(0, text=progress_text)
     # fill those as appropriate
-    #my_bar.progress(0.1, "Loading Database.  Please wait.")
-
-    #my_bar.progress(0.3, "Loading Model.  Please wait.")
-    
-    my_bar.progress(0.1, "Running RAG.  Please wait.")
+    my_bar.progress(0.3, "Loading Model.  Please wait.")
+    time.sleep(1) 
+    my_bar.progress(0.5, "Running RAG.  Please wait.")
     context = fetch_context(db, model, model_name, query, template)
 
     my_bar.progress(0.7, "Generating Answer.  Please wait.")
@@ -175,7 +173,7 @@ def generate_response(query,  model, template):
     my_bar.progress(0.9, "Post Processing.  Please wait.")
     
     my_bar.progress(1.0, "Done")
-    time. sleep(1) 
+    time.sleep(1) 
     my_bar.empty()  
     return response
 
@@ -204,9 +202,10 @@ def set_as_background_img(png_file):
 
 
 def stream_to_screen(response):
+    print(response)
     for word in response.split():
         yield word + " "
-        time.sleep(0.05)
+        time.sleep(0.15)
     
     
 if __name__=="__main__":
@@ -276,15 +275,12 @@ if __name__=="__main__":
                 Question: {question}\n> Context:\n>>>\n{context}\n>>>\nRelevant parts"""}
 
     # Loading and caching db and model
-    my_bar = st.progress(0, "Loading Database.  Please wait.")
-    my_bar.progress(0.1, "Loading Embedding & Database.  Please wait.")
+    #my_bar = st.progress(0, "Loading Database.  Please wait.")
+    #my_bar.progress(0.1, "Loading Embedding & Database.  Please wait.")
+    #my_bar.progress(0.7, "Loading Model.  Please wait.")
+
     db = load_db(device)
-    my_bar.progress(0.7, "Loading Model.  Please wait.")
     model, tokenizer = load_model(model_name)
-    my_bar.progress(1.0, "Done")
-    time. sleep(1) 
-    my_bar.empty()  
-
     
     response = False
     user_question = st.chat_input('What do you want to ask ..')
@@ -294,12 +290,12 @@ if __name__=="__main__":
     if user_question is not None and user_question!="":
         with st.chat_message("Human", avatar="🧔🏻"):
             st.write(user_question)
-            
+        """
         if response:
             with st.chat_message("AI", avatar="🏛️"):
                 # to empty response container after first pass
                 st.write(" ")
-            
+        """  
         response = generate_response(user_question,  model, all_templates)   
         with st.chat_message("AI", avatar="🏛️"):
             st.write_stream(stream_to_screen(response))