Update functions.py

functions.py

@@ -60,6 +60,7 @@ margin-bottom: 2.5rem">{}</div> """
 
 memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True, output_key='answer')
 
+
 #Stuff Chain Type Prompt template
 
 @st.cache_resource
@@ -89,12 +90,8 @@ def load_prompt():
     ----------------
     {context}"""
     
-    messages = [
-        SystemMessagePromptTemplate.from_template(system_template),
-        HumanMessagePromptTemplate.from_template("{question}")
-    ]
-    prompt = ChatPromptTemplate.from_messages(messages)
-
+    prompt = SystemMessagePromptTemplate.from_template(system_template),
+        
     return prompt
 
 ###################### Functions #######################################################################################
@@ -216,12 +213,10 @@ def gen_embeddings(embedding_model):
 
     return embeddings
     
-@st.cache_data
 def embed_text(query,embedding_model,_docsearch):
     
     '''Embed text and generate semantic search scores'''
-
-    chat_history = []
+    
     # llm = OpenAI(temperature=0)
     chat_llm = ChatOpenAI(streaming=True,
                           model_name = 'gpt-4',
@@ -235,10 +230,12 @@ def embed_text(query,embedding_model,_docsearch):
     #                              return_source_documents=True)
     
     chain = ConversationalRetrievalChain.from_llm(chat_llm, 
-                                                  retriever= _docsearch.as_retriever(),
-                                                  #qa_prompt = load_prompt(),
+                                                  retriever= _docsearch.as_retriever(search_kwargs={"k": 3})),
+                                                  get_chat_history=lambda h : h,
                                                   memory = memory,
                                                   return_source_documents=True)
+
+    chain.combine_docs_chain.llm_chain.prompt.messages[0] = load_prompt()
     
     answer = chain({"question": query})