app.py

# app.py
import os
import gradio as gr

from langchain.vectorstores import FAISS
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.document_loaders import TextLoader
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain.chains import RetrievalQA
from langchain.llms import HuggingFacePipeline
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline

# Optional: Set HF Token if needed
# os.environ['HUGGINGFACEHUB_API_TOKEN'] = 'hf_XXXX'

# Initialize embedding model
embedding_model = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")

# Load HF model (lightweight for CPU)
model_name = "google/flan-t5-small"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

# Wrap in pipeline
pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer, max_length=512)
llm = HuggingFacePipeline(pipeline=pipe)

def process_file(file_path):
    # Load & split document
    loader = TextLoader(file_path)
    documents = loader.load()
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
    docs = text_splitter.split_documents(documents)

    # Create vector DB
    vector_db = FAISS.from_documents(docs, embedding_model)
    retriever = vector_db.as_retriever()

    # Setup RetrievalQA chain
    qa_chain = RetrievalQA.from_chain_type(
        llm=llm,
        chain_type="stuff",
        retriever=retriever
    )

    return qa_chain

# Store the QA chain globally (across UI events)
qa_chain = None

def upload_and_prepare(file):
    global qa_chain
    # qa_chain = process_file(file)
    qa_chain = process_file(file.name)
    return "✅ Document processed. You can now ask questions!"

def ask_question(query):
    if not qa_chain:
        return "❌ Please upload a document first."
    response = qa_chain.invoke({"query": query})
    return response["result"]

# Gradio UI
with gr.Blocks() as demo:
    gr.Markdown("## 🧠 Ask Questions About Your Document (LangChain + Hugging Face)")
    
    with gr.Row():
        file_input = gr.File(label="📄 Upload .txt File", type="filepath")
        upload_btn = gr.Button("🔄 Process Document")

    upload_output = gr.Textbox(label="📁 Status", interactive=False)

    with gr.Row():
        query_input = gr.Textbox(label="❓ Your Question")
        query_btn = gr.Button("🧠 Get Answer")

    answer_output = gr.Textbox(label="✅ Answer", lines=4)

    upload_btn.click(upload_and_prepare, inputs=file_input, outputs=upload_output)
    query_btn.click(ask_question, inputs=query_input, outputs=answer_output)

# For local dev use: demo.launch()
# For HF Spaces
if __name__ == "__main__":
    demo.launch()