File size: 5,338 Bytes
7a0e422 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 |
import streamlit as st
import pinecone
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate
from langchain.vectorstores import Pinecone
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.chains.question_answering import load_qa_chain
from langchain.memory import ConversationBufferMemory, CombinedMemory, ConversationKGMemory
if "generated" not in st.session_state:
st.session_state["generated"] = []
if "past" not in st.session_state:
st.session_state["past"] = []
if "input" not in st.session_state:
st.session_state["input"] = ""
if "stored_session" not in st.session_state:
st.session_state["stored_session"] = []
# Define function to get user input
def get_text():
"""
Get user input text.
Returns:
str: The text entered by the user.
"""
input_text = st.text_input("You:", st.session_state["input"], key="input",
placeholder="Enter your message here...", label_visibility='hidden')
return input_text
def reset_entity_memory():
"""
Resets the entity memory to its initial state.
"""
st.session_state["entity_memory"] = CombinedMemory(memories=[KG, CBM])
def new_chat():
"""
Clears session state and starts a new chat.
"""
save = []
for i in range(len(st.session_state['generated'])-1, -1, -1):
save.append("User:" + st.session_state["past"][i])
save.append("Bot:" + st.session_state["generated"][i]['output_text'])
st.session_state["stored_session"].append(save)
st.session_state["generated"] = []
st.session_state["past"] = []
st.session_state["input"] = ""
if "entity_memory" in st.session_state:
reset_entity_memory()
# Set up the Streamlit app layout
st.title("ChatBot with Pinecone")
st.markdown(
'''
> :black[**A Chat Bot that queries your own corpus in Pinecone.**]
''')
# st.markdown(" > Powered by - 🦜 LangChain + OpenAI + Streamlit")
openai_api = st.sidebar.text_input("OpenAI API Key", type="password")
pinecone_api = st.sidebar.text_input("Pinecone API Key", type="password")
pinecone_env = st.sidebar.text_input("Pinecone Environment")
pinecone_index = st.sidebar.text_input("Pinecone Index")
MODEL = st.sidebar.selectbox("Model", ["gpt-4","gpt-3.5-turbo", "text-davinci-003"])
if openai_api and pinecone_api and pinecone_env and pinecone_index:
# Create Pinecone Instance
pinecone.init(api_key=pinecone_api, environment=pinecone_env)
# Create OpenAI Instance
llm = OpenAI(
temperature=0,
openai_api_key=openai_api,
model_name=MODEL,
max_tokens=2500,
)
# Create a ConversationEntityMemory object if not already created
if 'entity_memory' not in st.session_state:
KG = ConversationKGMemory(llm=llm, input_key="human_input")
CBM = ConversationBufferMemory(memory_key="chat_history", input_key="human_input")
st.session_state["entity_memory"] = CombinedMemory(memories=[KG, CBM])
# Set Template
# template = """You are a chatbot having a conversation with a human.
# Given the following extracted parts of a long document and a question, create a final answer.
# {context}
# Human: {human_input}
# Chatbot:"""
template = """As an advanced AI chatbot, you are engaged in a detailed conversation with a human user. You have access to a vast knowledge base stored in Pinecone's vector database. Your task is to leverage this information to provide comprehensive, accurate, and helpful responses. Given the following segments extracted from an extensive document and a user's question, analyze the context, draw specific information from the Pinecone memory, and formulate a well-rounded answer that not only addresses the query but also provides additional relevant information that could be beneficial to the user.
Context: {context}
Human: {human_input}
Your Response as Chatbot:"""
# Set the Prompt
prompt = PromptTemplate(
input_variables=["human_input", "context"],
template=template
)
# Get Context
embeddings = OpenAIEmbeddings(openai_api_key=openai_api)
docsearch = Pinecone.from_existing_index(pinecone_index, embedding=embeddings)
# Get user input
else:
st.error("Please enter your API keys in the sidebar.")
input_text = get_text()
if input_text:
# Fetch docs using user input for cosine similarity
docs = docsearch.similarity_search(input_text, k=6)
# # Get Response
chain = load_qa_chain(OpenAI(temperature=0, openai_api_key=openai_api), chain_type="stuff", memory=st.session_state["entity_memory"], prompt=prompt, verbose=True)
# Generate the output using user input and store it in the session state
output = chain({"input_documents": docs, "human_input": input_text}, return_only_outputs=True)
st.session_state.past.append(input_text)
st.session_state.generated.append(output)
with st.expander("Conversation"):
for i in range(len(st.session_state["generated"])-1, -1, -1):
st.info(st.session_state["past"][i])
st.success(st.session_state["generated"][i]['output_text'])
# # Create button to start a new chat
# st.sidebar.button("New Chat", on_click=new_chat, type="primary") |