from flask import Flask, render_template, request, jsonify
from langchain.vectorstores import FAISS
from langchain.chat_models import ChatOpenAI
from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.chains import ConversationalRetrievalChain
import os
from PyPDF2 import PdfReader
app = Flask(name)
Load PDF and extract text
pdf_file_obj = open(“docs/jat.pdf”, “rb”)
pdf_reader = PdfReader(pdf_file_obj)
num_pages = len(pdf_reader.pages)
detected_text = “”
for page_num in range(num_pages):
page_obj = pdf_reader.pages[page_num]
detected_text += page_obj.extract_text() + “\n\n”
pdf_file_obj.close()
Set OpenAI API Key
os.environ[“OPENAI_API_KEY”] = “”
Split text into documents
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
texts = text_splitter.create_documents([detected_text])
Build vector index
directory = “index_store”
vector_index = FAISS.from_documents(texts, OpenAIEmbeddings())
vector_index.save_local(directory)
Load vector index and set up retriever
vector_index = FAISS.load_local(directory, OpenAIEmbeddings())
retriever = vector_index.as_retriever(search_type=“similarity”, search_kwargs={“k”: 6})
Create ConversationalRetrievalChain
conv_interface = ConversationalRetrievalChain.from_llm(ChatOpenAI(temperature=0), retriever=retriever)
Store chat history
chat_history = # We won’t use this
@app.route(‘/’)
def index():
return render_template(‘chat.html’, user_input=“”, bot_response=“”)
@app.route(‘/api/chat’, methods=[‘POST’])
def chat():
user_input = request.form[‘user_input’]
# Get the bot's response
bot_response = conv_interface({"question": user_input, "chat_history": chat_history})["answer"]
return jsonify({"bot_response": bot_response})
if name == ‘main’:
app.run(debug=True, port=5001)
i want this code give me response only regards the my knowledge base. I just want to give results regards my docs only.