Spaces:
Sleeping
Sleeping
Upload 2 files
Browse files- app.py +76 -0
- requirements.txt +8 -0
app.py
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import os
|
| 2 |
+
import io
|
| 3 |
+
import streamlit as st
|
| 4 |
+
from PyPDF2 import PdfReader
|
| 5 |
+
from dotenv import load_dotenv
|
| 6 |
+
from langchain.embeddings.openai import OpenAIEmbeddings
|
| 7 |
+
from langchain.text_splitter import CharacterTextSplitter
|
| 8 |
+
from langchain.vectorstores import FAISS
|
| 9 |
+
# from constants import api_key
|
| 10 |
+
from langchain.chains.question_answering import load_qa_chain
|
| 11 |
+
from langchain.llms import OpenAI
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
def main():
|
| 15 |
+
# set the openai enviroment using api_key
|
| 16 |
+
load_dotenv()
|
| 17 |
+
|
| 18 |
+
st.title("PDF Reader")
|
| 19 |
+
|
| 20 |
+
# File uploader
|
| 21 |
+
uploaded_file = st.file_uploader("Upload a PDF file", type=["pdf"])
|
| 22 |
+
|
| 23 |
+
if uploaded_file is not None:
|
| 24 |
+
st.write("PDF uploaded successfully!")
|
| 25 |
+
|
| 26 |
+
# Read PDF contents
|
| 27 |
+
pdf_contents = uploaded_file.read()
|
| 28 |
+
|
| 29 |
+
pdf_file_like_object = io.BytesIO(pdf_contents)
|
| 30 |
+
pdfreader = PdfReader(pdf_file_like_object)
|
| 31 |
+
|
| 32 |
+
from typing_extensions import Concatenate
|
| 33 |
+
# read text from pdf
|
| 34 |
+
raw_text = ''
|
| 35 |
+
for i, page in enumerate(pdfreader.pages):
|
| 36 |
+
content = page.extract_text()
|
| 37 |
+
if content:
|
| 38 |
+
raw_text += content
|
| 39 |
+
|
| 40 |
+
# We need to split the text using Character Text Split such that it sshould not increse token size
|
| 41 |
+
text_splitter = CharacterTextSplitter(
|
| 42 |
+
separator = "\n",
|
| 43 |
+
chunk_size = 800,
|
| 44 |
+
chunk_overlap = 200,
|
| 45 |
+
length_function = len,
|
| 46 |
+
)
|
| 47 |
+
texts = text_splitter.split_text(raw_text)
|
| 48 |
+
|
| 49 |
+
embeddings = OpenAIEmbeddings()
|
| 50 |
+
document_search = FAISS.from_texts(texts, embeddings)
|
| 51 |
+
|
| 52 |
+
chain = load_qa_chain(OpenAI(), chain_type="stuff")
|
| 53 |
+
|
| 54 |
+
def bot(query, i):
|
| 55 |
+
if query:
|
| 56 |
+
docs = document_search.similarity_search(query)
|
| 57 |
+
response = chain.run(input_documents=docs, question=query)
|
| 58 |
+
st.text_area("Chatbot Response:", value=response, key=f"response_ker_{i}")
|
| 59 |
+
|
| 60 |
+
next_query = st.text_input("Enter your next question about the PDF : ", key=i+1)
|
| 61 |
+
|
| 62 |
+
i += 1
|
| 63 |
+
bot(query=next_query, i=i)
|
| 64 |
+
|
| 65 |
+
i = 0
|
| 66 |
+
query = st.text_input("Enter your question about the PDF : ", key=i)
|
| 67 |
+
bot(query=query, i=i)
|
| 68 |
+
|
| 69 |
+
|
| 70 |
+
i = 0
|
| 71 |
+
|
| 72 |
+
# # Display number of pages
|
| 73 |
+
st.write(f"Number of pages in PDF: {len(pdfreader.pages)}")
|
| 74 |
+
|
| 75 |
+
if __name__ == "__main__":
|
| 76 |
+
main()
|
requirements.txt
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
langchain==0.1.4
|
| 2 |
+
langchain_community==0.0.16
|
| 3 |
+
langchain_core==0.1.17
|
| 4 |
+
langchain_openai==0.0.5
|
| 5 |
+
streamlit==1.32.0
|
| 6 |
+
python-dotenv==1.0.1
|
| 7 |
+
PyPDF2==2.12.0
|
| 8 |
+
openai==1.14.0
|