app.py

import gradio as gr
import bs4
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import WebBaseLoader
from langchain_community.vectorstores import FAISS
#from langchain_community.vectorstores import Chroma
from langchain_community.embeddings import HuggingFaceBgeEmbeddings
#from langchain_community.embeddings import OllamaEmbeddings
import ollama

# Function to load, split, and retrieve documents from a URL
def load_and_retrieve_docs(url):
    loader = WebBaseLoader(
        web_paths=(url,),
        bs_kwargs=dict() 
    )
    docs = loader.load()
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
    splits = text_splitter.split_documents(docs)
    embeddings = HuggingFaceBgeEmbeddings(model_name="BAAI/bge-small-en-v1.5",
                                          model_kwargs={'device':'cpu'},
                                          encode_kwargs={'normalize_embeddings':True})
    vectorstore = FAISS.from_documents(documents=splits, embedding=embeddings)
    return vectorstore.as_retriever()

# Function to initialize vector embedding with FAISS vector store
def vector_embedding():
    if "vectors" not in st.session_state:
        st.session_state.embeddings = HuggingFaceBgeEmbeddings(model_name="BAAI/bge-small-en-v1.5",
                                                               model_kwargs={'device':'cpu'},
                                                               encode_kwargs={'normalize_embeddings':True})
        st.session_state.loader = PyPDFDirectoryLoader("./Data_Science")  # Data Ingestion
        st.session_state.docs = st.session_state.loader.load()  # Document Loading
        st.session_state.text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)  # Chunk Creation
        st.session_state.final_documents = st.session_state.text_splitter.split_documents(st.session_state.docs[:20])  # Splitting
        st.session_state.vectors = FAISS.from_documents(st.session_state.final_documents, st.session_state.embeddings)  # Vector HuggingFace embeddings
        st.write("Vector Store DB Is Ready")
    else:
        st.write("Vectors already initialized.")

# Function to format documents
def format_docs(docs):
    return "\n\n".join(doc.page_content for doc in docs)

# Function that defines the RAG chain
def rag_chain(url, question):
    retriever = load_and_retrieve_docs(url)
    retrieved_docs = retriever.invoke(question)
    formatted_context = format_docs(retrieved_docs)
    formatted_prompt = f"Question: {question}\n\nContext: {formatted_context}"
    response = ollama.chat(model='llama3', messages=[{'role': 'user', 'content': formatted_prompt}])
    return response['message']['content']

# Gradio interface
iface = gr.Interface(
    fn=rag_chain,
    inputs=["text", "text"],
    outputs="text",
    title="Rocky Bot",
    description="Enter a URL and a query to get answers from the RAG chain."
)

# Launch the app
iface.launch()