Update app.py

app.py

@@ -1,93 +1,40 @@
 import streamlit as st
+import pandas as pd
 from langchain_core.messages import AIMessage, HumanMessage
-from langchain_openai import ChatOpenAI
-
-from langchain_core.output_parsers import StrOutputParser
-from langchain_core.prompts import ChatPromptTemplate
 
-# from langchain.chat_models import AzureChatOpenAI
-from langchain_openai import AzureChatOpenAI
-from langchain.schema import HumanMessage, SystemMessage
-from langchain_core.prompts.chat import SystemMessagePromptTemplate, HumanMessagePromptTemplate, ChatPromptTemplate
-from azure_openai import qt
+from azure_openai import qt, get_response
 from retriver import search_and_reconstruct
-# Initialize an instance of AzureOpenAI using the specified settings
-import pandas as pd
-
-
-# LLM Langchain Definition
-OPENAI_API_KEY = st.secrets['azure_api_key']
-OPENAI_API_TYPE = "azure"
-OPENAI_API_BASE = "https://davidfearn-gpt4.openai.azure.com"
-# OPENAI_API_VERSION = "2024-02-01"
-OPENAI_API_VERSION = "2024-08-01-preview"
-# OPENAI_MODEL = "gpt4-turbo-1106"
-OPENAI_MODEL = "gpt-4o"
-# Initialize an instance of AzureOpenAI using the specified settings
 
+placeHolderPersonaQTE = "{assetGlossary}, {conversationToDate}"
+placeHolderPersonaRAGSystem = "{assetGlossary}"
+placeHolderPersonaRAGUser = "{query}, {knowledge}"
 
-def get_response(chat_history, qte, knowledge, temp1, temp2, tokens1, tokens2, persona2SystemMessage, persona2UserMessage):
-
-    llm = AzureChatOpenAI(
-        openai_api_version=OPENAI_API_VERSION,
-        openai_api_key=OPENAI_API_KEY,
-        azure_endpoint=OPENAI_API_BASE,
-        openai_api_type=OPENAI_API_TYPE,
-        deployment_name=OPENAI_MODEL,
-        temperature=temp2, 
-        max_tokens=tokens2
-            # Name of the deployment for identification
-    )
-
-
-    system_message_template = SystemMessagePromptTemplate.from_template(persona2SystemMessage)
-    # human_message_template = HumanMessagePromptTemplate.from_template(persona2UserMessage)
-    
-    # Create a chat prompt template combining system and human messages
-    # prompt = ChatPromptTemplate.from_messages([system_message_template, human_message_template])
-    prompt = ChatPromptTemplate.from_messages([system_message_template, persona2UserMessage])
-
-    chain = prompt | llm | StrOutputParser()
-    
-    return chain.stream({
-        "query": chat_history,
-        "knowledge": knowledge
-    })
 
-placeHolderPersona1system = "place holder"
-placeHolderPersona2user = "Query: {query}, Knowledge: {knowledge}"
 # app config
 st.set_page_config(page_title="Reg Intel Chatbot", page_icon="🤖")
 st.title("Reg Intel Toolbox :toolbox:")
 
 # Sidebar for inputting personas
 st.sidebar.title("RAG System Designer")
-# st.sidebar.subheader("Welcome Message")
-# welcomeMessage = st.sidebar.text_area("Define Intake Persona", value=welcomeMessage, height=300)
 st.sidebar.header("Query Designer Config")
-# numberOfQuestions = st.sidebar.slider("Number of Questions", min_value=0, max_value=10, step=1, value=5, key='persona1_questions')
-persona1SystemMessage = st.sidebar.text_area("Query Designer System Message", value=placeHolderPersona1system, height=300)
+persona1SystemMessage = st.sidebar.text_area("Query Designer System Message", value=placeHolderPersonaQTE, height=300)
 temp1 = st.sidebar.slider("Temperature", min_value=0.0, max_value=1.0, step=0.1, value=0.6, key='persona1_temp')
 tokens1 = st.sidebar.slider("Tokens", min_value=0, max_value=4000, step=100, value=500, key='persona1_tokens')
 st.sidebar.subheader("Number of Search Results")
 k = st.sidebar.slider("Returned Docs", min_value=1, max_value=10, step=1, value=3, key='k')
-pagesReturned = st.sidebar.slider("Number of Pages Returned", min_value=1, max_value=10, step=1, value=1, key='pagesReturned')
-
-
+pagesReturned = st.sidebar.slider("Number of Returned Pages", min_value=1, max_value=10, step=1, value=1, key='pagesReturned')
 st.sidebar.header("Engineered Prompt Config")
-persona2SystemMessage = st.sidebar.text_area("Answer Creation System Message", value=placeHolderPersona1system, height=300)
-persona2UserMessage = st.sidebar.text_area("Answer Creation User Message", value=placeHolderPersona2user, height=300)
+persona2SystemMessage = st.sidebar.text_area("Answer Creation System Message", value=placeHolderPersonaRAGSystem, height=300)
+persona2UserMessage = st.sidebar.text_area("Answer Creation User Message", value=placeHolderPersonaRAGUser, height=300)
 temp2 = st.sidebar.slider("Temperature", min_value=0.0, max_value=1.0, step=0.1, value=0.6, key='persona2_temp')
 tokens2 = st.sidebar.slider("Tokens", min_value=0, max_value=4000, step=100, value=500, key='persona2_tokens')
 
-
 # session state
 if "chat_history" not in st.session_state:
     st.session_state.chat_history = [
         AIMessage(content="Hello, I am the GSK Reg Intel Assistant. How can I help you?"),
     ]
 
-
 # conversation
 for message in st.session_state.chat_history:
     if isinstance(message, AIMessage):
@@ -97,7 +44,6 @@ for message in st.session_state.chat_history:
         with st.chat_message("Human"):
             st.write(message.content)
 
-
 # user input
 user_query = st.chat_input("Type your message here...")
 if user_query is not None and user_query != "":
@@ -106,15 +52,12 @@ if user_query is not None and user_query != "":
     with st.chat_message("Human"):
         st.markdown(user_query)
 
-    with st.chat_message("AI"):
-        qte = qt(persona1SystemMessage, st.session_state.chat_history, temp1, tokens1)
-        knowledge = search_and_reconstruct(qte, k, pagesReturned)
-        response = st.write_stream(get_response(st.session_state.chat_history, qte, knowledge, temp1, temp2, tokens1, tokens2, persona2SystemMessage, persona2UserMessage))
-
-    st.session_state.chat_history.append(AIMessage(content=response))
+   
+    qte = qt(persona1SystemMessage, st.session_state.chat_history, temp1, tokens1, "GSKGlossary")
     st.sidebar.header("QTE and Knowledge Results")
     st.sidebar.header("QTE")
     st.sidebar.text(qte)
+    knowledge = search_and_reconstruct(qte, k, pagesReturned)
     
     if knowledge:
         
@@ -134,4 +77,8 @@ if user_query is not None and user_query != "":
         st.sidebar.write("### Knowledge Base Results")
         st.sidebar.dataframe(df)  # Adjust height as needed
     else:
-        st.sidebar.write("No relevant knowledge base results found.")
+        st.sidebar.write("No relevant knowledge base results found.")
+    
+    with st.chat_message("AI"):
+        response = st.write_stream(get_response(st.session_state.chat_history, qte, knowledge, temp1, temp2, tokens1, tokens2, persona2SystemMessage, persona2UserMessage, "GSKGlossary"))
+        st.session_state.chat_history.append(AIMessage(content=response))