Spaces:
Sleeping
Sleeping
| from PyPDF2 import PdfReader | |
| from langchain.text_splitter import RecursiveCharacterTextSplitter | |
| from langchain.vectorstores import FAISS | |
| from transformers import pipeline | |
| import streamlit as st | |
| import requests | |
| from io import BytesIO | |
| # Set up Hugging Face model pipeline for text generation | |
| pipe = pipeline("text-generation", model="meta-llama/Llama-Guard-3-8B-INT8") | |
| # List of GitHub PDF URLs | |
| PDF_URLS = [ | |
| "https://github.com/TahirSher/GenAI_Lawyers_Guide/blob/main/bi%20pat%20graphs.pdf", | |
| "https://github.com/TahirSher/GenAI_Lawyers_Guide/blob/main/bi-partite.pdf", | |
| # Add more document links as needed | |
| ] | |
| def fetch_pdf_text_from_github(urls): | |
| text = "" | |
| for url in urls: | |
| response = requests.get(url) | |
| if response.status_code == 200: | |
| pdf_file = BytesIO(response.content) | |
| pdf_reader = PdfReader(pdf_file) | |
| for page in pdf_reader.pages: | |
| page_text = page.extract_text() | |
| if page_text: | |
| text += page_text | |
| else: | |
| st.error(f"Failed to fetch PDF from URL: {url}") | |
| return text | |
| def get_text_chunks(text): | |
| text_splitter = RecursiveCharacterTextSplitter(chunk_size=10000, chunk_overlap=1000) | |
| chunks = text_splitter.split_text(text) | |
| return chunks | |
| def load_or_create_vector_store(text_chunks): | |
| embeddings = FAISS.get_default_embeddings() | |
| vector_store = FAISS.from_texts(text_chunks, embedding=embeddings) | |
| return vector_store | |
| def generate_answer(user_question, context_text): | |
| # Format the input message for the pipeline | |
| messages = [ | |
| {"role": "user", "content": f"Context: {context_text}\nQuestion: {user_question}"} | |
| ] | |
| # Generate response using the pipeline | |
| response = pipe(messages, max_length=250, do_sample=True) | |
| return response[0]['generated_text'][:250] # Limit response to 250 characters | |
| def user_input(user_question, vector_store): | |
| docs = vector_store.similarity_search(user_question) | |
| context_text = " ".join([doc.page_content for doc in docs]) | |
| return generate_answer(user_question, context_text) | |
| def main(): | |
| st.set_page_config(page_title="RAG-based PDF Chat", layout="centered", page_icon="📄") | |
| st.title("📄 Query PDF Documents on GitHub") | |
| # Load documents from GitHub | |
| raw_text = fetch_pdf_text_from_github(PDF_URLS) | |
| text_chunks = get_text_chunks(raw_text) | |
| vector_store = load_or_create_vector_store(text_chunks) | |
| # User question input | |
| user_question = st.text_input("Ask a Question:", placeholder="Type your question here...") | |
| if st.button("Get Response"): | |
| if not user_question: | |
| st.warning("Please enter a question before submitting.") | |
| else: | |
| with st.spinner("Generating response..."): | |
| answer = user_input(user_question, vector_store) | |
| st.markdown(f"**🤖 AI:** {answer}") | |
| if __name__ == "__main__": | |
| main() | |