Spaces:
Sleeping
Sleeping
File size: 2,281 Bytes
fa3ff28 |
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 |
import tempfile
import streamlit as st
import pickle
from langchain_google_genai import GoogleGenerativeAIEmbeddings
from langchain_community.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.vectorstores import FAISS
import faiss
def process_pdf(uploaded_file):
all_documents = []
st.session_state.embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001")
main_placeholder = st.empty()
# Creating a temporary file to store the uploaded PDF's
main_placeholder.text("Data Loading...Started...✅✅✅")
for uploaded_file in uploaded_file:
with tempfile.NamedTemporaryFile(delete=False , suffix='.pdf') as temp_file:
temp_file.write(uploaded_file.read()) ## write file to temporary
temp_file_path = temp_file.name # Get the temporary file path
# Load the PDF's from the temporary file path
loader = PyPDFLoader(temp_file_path) # Document loader
doc= loader.load() # load Document
main_placeholder.text("Text Splitter...Started...✅✅✅")
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200) # Recursive Character String
#final_documents = text_splitter.split_documents(doc)# splitting
final_documents = text_splitter.split_documents(doc)
all_documents.extend(final_documents)
if all_documents:
main_placeholder.text("Embedding Vector Started Building...✅✅✅")
st.session_state.vectors = FAISS.from_documents(all_documents,st.session_state.embeddings)
st.session_state.docs = all_documents
# Save FAISS vector store to disk
faiss_index = st.session_state.vectors.index # Extract FAISS index
faiss.write_index(faiss_index, "faiss_index.bin") # Save index to a binary file
main_placeholder.text("Vector database created!...✅✅✅")
else:
st.error("No documents found after processing the uploaded files or the pdf is corrupted / unsupported.")
|