Spaces:
Runtime error
Runtime error
#!/usr/bin/env python | |
# coding: utf-8 | |
# In[1]: | |
#import necessary packages | |
import os | |
from openai import AsyncOpenAI # importing openai for API usage | |
import chainlit as cl # importing chainlit for our app | |
from chainlit.playground.providers import ChatOpenAI # importing ChatOpenAI tools | |
from langchain_community.vectorstores import FAISS | |
from langchain_openai import OpenAIEmbeddings | |
from langchain.prompts import ChatPromptTemplate | |
from operator import itemgetter | |
from langchain_core.runnables import RunnablePassthrough | |
from langchain_community.vectorstores import FAISS | |
from langchain_openai import ChatOpenAI | |
from langchain.retrievers import MultiQueryRetriever | |
from langchain.chains.combine_documents import create_stuff_documents_chain | |
from langchain.chains import create_retrieval_chain | |
from langchain import hub | |
#from langchain.utils import itemgetter, RunnablePassthrough | |
#from langchain.chains import build_chain | |
#from langchain.text_splitter import RecursiveCharacterTextSplitter | |
#from langchain_community.document_loaders import PyMuPDFLoader | |
# In[2]: | |
#load environment var | |
from dotenv import load_dotenv | |
load_dotenv() | |
# In[3]: | |
#load in embeddings model | |
out_fp = './data' | |
embeddings = OpenAIEmbeddings(model="text-embedding-3-small") | |
#vector_store = FAISS.from_documents(documents, embeddings) | |
faiss_fn = 'nvidia_10k_faiss_index.bin' | |
vector_store=FAISS.load_local(out_fp+faiss_fn, embeddings, allow_dangerous_deserialization=True) | |
retriever = vector_store.as_retriever() | |
openai_llm = ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0) | |
# In[4]: | |
# ChatOpenAI Templates | |
template = """Answer the question based only on the following context. If you cannot answer the question with the context, respond with 'I don't know'. You'll get a big bonus and a potential promotion if you provide a high quality answer: | |
Context: | |
{context} | |
Question: | |
{question} | |
""" | |
prompt_template = ChatPromptTemplate.from_template(template) | |
# In[5]: | |
#create chain | |
retrieval_qa_prompt = hub.pull("langchain-ai/retrieval-qa-chat") | |
primary_qa_llm = ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0) | |
advanced_retriever = MultiQueryRetriever.from_llm(retriever=retriever, llm=primary_qa_llm) | |
document_chain = create_stuff_documents_chain(primary_qa_llm, retrieval_qa_prompt) | |
retrieval_chain = create_retrieval_chain(advanced_retriever, document_chain) | |
# In[6]: | |
# marks a function that will be executed at the start of a user session | |
async def start_chat(): | |
settings = { | |
"model": "gpt-3.5-turbo", | |
"temperature": 0, | |
"max_tokens": 250, | |
"top_p": 1, | |
"frequency_penalty": 0, | |
"presence_penalty": 0, | |
} | |
cl.user_session.set("settings", settings) | |
# In[8]: | |
# marks a function that should be run each time the chatbot receives a message from a user | |
async def main(message: cl.Message): | |
settings = cl.user_session.get("settings") | |
# Use the retrieval_augmented_qa_chain_openai pipeline with the user's question | |
question = message.content # Extracting the question from the message content | |
response = retrieval_chain.invoke({"input": question}) # Invoke the pipeline | |
#print(response['answer']) | |
# Extract the response content and context documents | |
response_content = response['answer'] | |
#context_documents = '\n'.join([document.page_content for document in response["context"]]) | |
#page_numbers = set([document.metadata['page'] for document in response["context"]]) | |
# Stream the response content back to the user | |
msg = cl.Message(content="") | |
await msg.stream_token(response_content) | |
# In[ ]: | |