Spaces:
Sleeping
Sleeping
Test
Browse files- app.py +26 -0
- requirements.txt +10 -0
- utils/PDF_Reader.py +31 -0
- utils/QA_Bot.py +37 -0
- utils/QnA.py +12 -0
- utils/__pycache__/Api_Key.cpython-311.pyc +0 -0
- utils/__pycache__/PDF_Reader.cpython-311.pyc +0 -0
- utils/__pycache__/QA_Bot.cpython-311.pyc +0 -0
- utils/__pycache__/QnA.cpython-311.pyc +0 -0
app.py
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import ollama
|
3 |
+
from utils.QA_Bot import QA_Bot
|
4 |
+
from utils.PDF_Reader import PDF_4_QA
|
5 |
+
|
6 |
+
ollama.pull("llama3.1")
|
7 |
+
|
8 |
+
# Streamlit app
|
9 |
+
def main():
|
10 |
+
st.sidebar.title("Upload PDF")
|
11 |
+
st.sidebar.write("Download Demo PDF file from Below....")
|
12 |
+
with open("Demo_Document/Kia_EV6.pdf", "rb") as file:
|
13 |
+
btn = st.sidebar.download_button(
|
14 |
+
label="Download PDF",
|
15 |
+
data=file,
|
16 |
+
file_name="Kia_EV6.pdf"
|
17 |
+
)
|
18 |
+
|
19 |
+
uploaded_file = st.sidebar.file_uploader("Choose a PDF file", type="pdf")
|
20 |
+
if uploaded_file is not None:
|
21 |
+
st.sidebar.success("File uploaded successfully.")
|
22 |
+
vector_store = PDF_4_QA(uploaded_file)
|
23 |
+
QA_Bot(vector_store)
|
24 |
+
|
25 |
+
if __name__ == '__main__':
|
26 |
+
main()
|
requirements.txt
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
langchain==0.3.17
|
2 |
+
langchain_ollama==0.2.3
|
3 |
+
ollama==0.4.7
|
4 |
+
PyPDF2==3.0.1
|
5 |
+
streamlit==1.29.0
|
6 |
+
streamlit==1.32.2
|
7 |
+
streamlit==1.35.0
|
8 |
+
streamlit==1.26.0
|
9 |
+
streamlit==1.41.1
|
10 |
+
~treamlit==1.29.0
|
utils/PDF_Reader.py
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import PyPDF2
|
2 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
3 |
+
from langchain.embeddings import HuggingFaceBgeEmbeddings
|
4 |
+
from langchain.vectorstores import FAISS
|
5 |
+
|
6 |
+
def read_pdf(uploaded_file):
|
7 |
+
pdf_reader = PyPDF2.PdfReader(uploaded_file)
|
8 |
+
text = ""
|
9 |
+
for page in pdf_reader.pages:
|
10 |
+
text += page.extract_text()
|
11 |
+
return text
|
12 |
+
|
13 |
+
def Chunks(docs):
|
14 |
+
text_splitter = RecursiveCharacterTextSplitter(
|
15 |
+
# Set a really small chunk size, just to show.
|
16 |
+
chunk_size = 1000,
|
17 |
+
chunk_overlap = 100,
|
18 |
+
)
|
19 |
+
doc = text_splitter.split_text(docs)
|
20 |
+
return doc
|
21 |
+
|
22 |
+
|
23 |
+
def PDF_4_QA(file):
|
24 |
+
content = read_pdf(file)
|
25 |
+
pdf_chunks = Chunks(docs=content)
|
26 |
+
|
27 |
+
embeddings = HuggingFaceBgeEmbeddings(model_name='sentence-transformers/all-MiniLM-L6-v2',
|
28 |
+
model_kwargs={'device': 'cpu'})
|
29 |
+
vectorstore_openai = FAISS.from_texts(pdf_chunks, embeddings)
|
30 |
+
|
31 |
+
return vectorstore_openai
|
utils/QA_Bot.py
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
from utils.QnA import Q_A
|
3 |
+
import re,time
|
4 |
+
|
5 |
+
|
6 |
+
def QA_Bot(vectorstore):
|
7 |
+
st.title("Q&A Bot")
|
8 |
+
# Initialize chat history
|
9 |
+
if "messages" not in st.session_state:
|
10 |
+
st.session_state.messages = []
|
11 |
+
|
12 |
+
# Display chat messages from history on app rerun
|
13 |
+
for message in st.session_state.messages:
|
14 |
+
with st.chat_message(message["role"]):
|
15 |
+
st.markdown(message["content"])
|
16 |
+
|
17 |
+
# React to user input
|
18 |
+
if prompt := st.chat_input("What is up?"):
|
19 |
+
# Display user message in chat message container
|
20 |
+
st.chat_message("user").markdown(prompt)
|
21 |
+
# Add user message to chat history
|
22 |
+
st.session_state.messages.append({"role": "user", "content": prompt})
|
23 |
+
|
24 |
+
ai_response = Q_A(vectorstore,prompt)
|
25 |
+
response = f"Echo: {ai_response}"
|
26 |
+
# Display assistant response in chat message container
|
27 |
+
with st.chat_message("assistant"):
|
28 |
+
message_placeholder = st.empty()
|
29 |
+
full_response = ""
|
30 |
+
for chunk in re.split(r'(\s+)', response):
|
31 |
+
full_response += chunk + " "
|
32 |
+
time.sleep(0.01)
|
33 |
+
|
34 |
+
# Add a blinking cursor to simulate typing
|
35 |
+
message_placeholder.markdown(full_response + "▌")
|
36 |
+
# Add assistant response to chat history
|
37 |
+
st.session_state.messages.append({"role": "assistant", "content": full_response})
|
utils/QnA.py
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from langchain.chains import RetrievalQA
|
2 |
+
from langchain_ollama import ChatOllama
|
3 |
+
|
4 |
+
def Q_A(vectorstore,question):
|
5 |
+
ollama_llm = ChatOllama(
|
6 |
+
model="llama3.1",
|
7 |
+
temperature=0.5,
|
8 |
+
)
|
9 |
+
qa = RetrievalQA.from_chain_type(llm=ollama_llm, chain_type="stuff", retriever=vectorstore.as_retriever())
|
10 |
+
answer = qa.invoke(question)
|
11 |
+
|
12 |
+
return answer['result']
|
utils/__pycache__/Api_Key.cpython-311.pyc
ADDED
Binary file (285 Bytes). View file
|
|
utils/__pycache__/PDF_Reader.cpython-311.pyc
ADDED
Binary file (1.65 kB). View file
|
|
utils/__pycache__/QA_Bot.cpython-311.pyc
ADDED
Binary file (2.28 kB). View file
|
|
utils/__pycache__/QnA.cpython-311.pyc
ADDED
Binary file (848 Bytes). View file
|
|