Update app.py

app.py

@@ -4,11 +4,13 @@ from langchain.chat_models import ChatOpenAI
 from langchain.embeddings import OpenAIEmbeddings
 from langchain.vectorstores import FAISS
 from langchain.chains import RetrievalQA
+from langchain.chains import ConversationalRetrievalChain
 from langchain.prompts import PromptTemplate
 from langchain.document_loaders import WebBaseLoader, AsyncHtmlLoader
 from langchain.document_transformers import Html2TextTransformer
 from langchain.callbacks import get_openai_callback
 from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.memory import ConversationBufferWindowMemory
 
 import asyncio
 from langchain.docstore.document import Document
@@ -28,19 +30,33 @@ input_url = st.text_input("Inserisci url:")
 question = st.text_area("Chiedi pure:")
 
 #generate the main prompt
-prompt_template = """Use the following pieces of context to answer the question at the end. If you don't know the answer, just say that you don't know, don't try to make up an answer. 
-
+prompt_template = """Intruction: You are a website agent that is talking with a human. Use only the chat history and the following information:
 
 {context}
 
-Question: {question}
-Answer:
+to answer in a helpful manner to the question. If you don't know the answer - say that you don't know. 
+Keep your replies short, compassionate and informative. 
+{chat_history}
+
+### Input: {question}
+### Response:
 """
 
-PROMPT = PromptTemplate(
-    template=prompt_template, input_variables=["context", "question"]
+
+prompt = PromptTemplate(
+    template=prompt_template, input_variables=["context", "question", "chat_history"]
+)
+
+
+memory = ConversationBufferWindowMemory(
+    memory_key="chat_history",
+    ai_prefix="### Response",
+    human_prefix="### Input",
+    output_key="answer",
+    return_messages=True
 )
 
+
 if st.button("Invia", type="primary"):
     
     loader = AsyncHtmlLoader(input_url)
@@ -65,15 +81,26 @@ if st.button("Invia", type="primary"):
     retriever = vectordb.as_retriever(search_kwargs={"k": 3})
     llm = ChatOpenAI(model_name = "gpt-3.5-turbo", temperature=0)
     relevant_docs = retriever.get_relevant_documents(question)
-    qa = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=retriever, chain_type_kwargs={"prompt": PROMPT})
+
+
+    chain = ConversationalRetrievalChain.from_llm(
+        llm,
+        chain_type='stuff',
+        retriever=retriever,
+        memory=memory,
+        combine_docs_chain_kwargs={"prompt": prompt},
+        verbose=True
+    )
+
+    #qa = RetrievalQA.from_chain_type(llm=llm, chain_type="stuff", retriever=retriever, chain_type_kwargs={"prompt": prompt})
 
      # Write answer and sources
     answer = st.empty()
 
     with get_openai_callback() as cb:
         #run the chain and generate response
-        response = qa(question)
+        response = chain(question)
         print(cb)
     
-    answer.write(response['result'])
+    answer.write(response["answer"])
     st.write(relevant_docs)