Update handler.py

handler.py

@@ -87,19 +87,20 @@ class EndpointHandler():
         compressor = LLMChainExtractor.from_llm(chat)
         retriever = ContextualCompressionRetriever(base_compressor=compressor, base_retriever=retriever)
 
-        _template = """Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question, in its original language.
+        _template = """[INST] Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question, in its original language.
         Chat History:
         {chat_history}
         Follow Up Input: {question}
-        Standalone question:"""
+        Standalone question: [/INST]"""
         
         CONDENSE_QUESTION_PROMPT = PromptTemplate.from_template(_template)
         
-        template = """Answer the question based only on the following context:
+        template = """[INST] Answer the question based only on the following context:
         {context}
         
-        Question: {question}
+        Question: {question} [/INST]
         """
+        
         ANSWER_PROMPT = ChatPromptTemplate.from_template(template)
 
         self.memory = ConversationBufferMemory(