Update app.py

app.py

@@ -389,15 +389,15 @@ chain_neo4j = (
 #         return "Sorry, I encountered an error while processing your request.", []
 
 
-# Short Prompt Template for Phi-3.5 Proprietary Model
+# # Short Prompt Template for Phi-3.5 Proprietary Model
 
-phi_short_template = f"""
-As an expert on Birmingham, Alabama, I will provide concise, accurate, and informative responses to your queries based on 128 token limit . Given the sunny weather today, {current_date}, feel free to ask me anything you need to know about the city.
-Provide only the direct answer to the question without any follow-up questions.
-{{context}}
-Question: {{question}}
-Answer:
-"""
+# phi_short_template = f"""
+# As an expert on Birmingham, Alabama, I will provide concise, accurate, and informative responses to your queries based on 128 token limit . Given the sunny weather today, {current_date}, feel free to ask me anything you need to know about the city.
+# Provide only the direct answer to the question without any follow-up questions.
+# {{context}}
+# Question: {{question}}
+# Answer:
+# """
 
 
 import re
@@ -449,11 +449,12 @@ def generate_answer(message, choice, retrieval_mode, selected_model):
             elif selected_model == phi_pipe:
                 retriever = phi_retriever
                 context = retriever.get_relevant_documents(message)
-                prompt = phi_short_template.format(context=context, question=message)
-
-                logging.debug(f"Phi-3.5 Prompt: {prompt}")
+                
+                # Format the Phi-3.5 prompt according to the required format
+                phi_prompt = f"<|system|>\nYou are a helpful assistant.<|end|>\n<|user|>\n{message}\n<|end|>\n<|assistant|>\n"
+                logging.debug(f"Phi-3.5 Prompt: {phi_prompt}")
 
-                response = selected_model(prompt, **{
+                response = selected_model(phi_prompt, **{
                     "max_new_tokens": 128,  # Increased to handle longer responses
                     "return_full_text": False,
                     "temperature": 0.7,  # Adjusted to avoid cutting off