app.py

import streamlit as st
import pickle
import os
import torch
from tqdm.auto import tqdm
from langchain.text_splitter import RecursiveCharacterTextSplitter

# from langchain.vectorstores import Chroma
from langchain.vectorstores import FAISS
from langchain.embeddings import HuggingFaceInstructEmbeddings


from langchain import HuggingFacePipeline
from langchain.chains import RetrievalQA



st.set_page_config(
    page_title = 'aitGPT',
    page_icon = '✅')


st.markdown("# Hello")


with open("ait-web-document", "rb") as fp:
    ait_web_documents = pickle.load(fp)
    
    
text_splitter = RecursiveCharacterTextSplitter(
    # Set a really small chunk size, just to show.
    chunk_size = 500,
    chunk_overlap  = 100,
    length_function = len,
)

chunked_text = text_splitter.create_documents([doc for doc in tqdm(ait_web_documents)])


st.markdown(f"Number of Documents: {len(ait_web_documents)}")
st.markdown(f"Number of chunked texts: {len(chunked_text)}")


embedding_model = HuggingFaceInstructEmbeddings(model_name='hkunlp/instructor-base',
                                                model_kwargs = {'device': torch.device('cuda' if torch.cuda.is_available() else 'cpu')})

vector_database = FAISS.load_local("faiss_index", embedding_model)
print("load done")




query_input = st.text_input(label= 'your question')
def retrieve_document(query_input):
    related_doc = vector_database.similarity_search(query_input)
    return related_doc

output = st.text_area(label = "Here is the relevant documents",
                      value = retrieve_document(query_input))