Update functions.py

functions.py

@@ -30,10 +30,7 @@ from langchain.embeddings import HuggingFaceEmbeddings, HuggingFaceBgeEmbeddings
 from langchain.vectorstores import FAISS
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.chat_models import ChatOpenAI
-from langchain.callbacks import StdOutCallbackHandler
-from langchain.chains import ConversationalRetrievalChain, QAGenerationChain, LLMChain
-from langchain.memory import ConversationBufferMemory
-from langchain.chains.question_answering import load_qa_chain
+from langchain.chains import QAGenerationChain
 
 from langchain.callbacks import StreamlitCallbackHandler
 from langchain.agents import OpenAIFunctionsAgent, AgentExecutor
@@ -66,43 +63,6 @@ time_str = time.strftime("%d%m%Y-%H%M%S")
 HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; 
 margin-bottom: 2.5rem">{}</div> """
 
-memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True, output_key='answer')
-
-
-#Stuff Chain Type Prompt template
-
-@st.cache_data
-def load_prompt():
-
-    system_template="""Use only the following pieces of earnings context to answer the users question accurately.
-    Do not use any information not provided in the earnings context and remember you are a to speak like a finance expert.
-    If you don't know the answer, just say 'There is no relevant answer in the given earnings call transcript', 
-    don't try to make up an answer.
-    
-    ALWAYS return a "SOURCES" part in your answer.
-    The "SOURCES" part should be a reference to the source of the document from which you got your answer.
-    
-    Remember, do not reference any information not given in the context.
-
-    If the answer is not available in the given context just say 'There is no relevant answer in the given earnings call transcript'
-    
-    Follow the below format when answering:
-    
-    Question: {question}
-    SOURCES: [xyz]
-      
-    Begin!
-    ----------------
-    {context}"""
-    
-    messages = [
-        SystemMessagePromptTemplate.from_template(system_template),
-        HumanMessagePromptTemplate.from_template("{question}")
-    ]
-    prompt = ChatPromptTemplate.from_messages(messages)
-        
-    return prompt
-
 
 ###################### Functions #######################################################################################
 
@@ -597,7 +557,8 @@ def create_vectorstore(corpus, title, embedding_model, chunk_size=1000, overlap=
     vectorstore = FAISS.from_texts(texts, embeddings, metadatas=[{"source": i} for i in range(len(texts))])
 
     return vectorstore
-    
+
+@st.cache_data
 def create_memory_and_agent(query,_docsearch):
     
     '''Embed text and generate semantic search scores'''