app.py

import streamlit as st
import tempfile
import logging
from typing import List, Optional
import torch
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from langchain_community.llms import HuggingFacePipeline
from langchain.chains.summarize import load_summarize_chain
from langchain.schema import Document
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.prompts import PromptTemplate
from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Constants
EMBEDDING_MODEL = 'sentence-transformers/all-MiniLM-L6-v2'
DEFAULT_MODEL = "distilgpt2"
MAX_LENGTH_FRACTION = 0.2  # Set max_length to 20% of input length

# Check for GPU
device = "cuda" if torch.cuda.is_available() else "cpu"
st.sidebar.write(f"Using device: {device}")

@st.cache_data
def load_embeddings(model_name: str) -> Optional[HuggingFaceEmbeddings]:
    """Load the embedding model."""
    try:
        return HuggingFaceEmbeddings(model_name=model_name)
    except Exception as e:
        logger.error(f"Failed to load embeddings: {e}")
        return None

@st.cache_data
def load_llm(model_name: str, max_length: int) -> Optional[HuggingFacePipeline]:
    """Load the language model."""
    try:
        tokenizer = AutoTokenizer.from_pretrained(model_name)
        model = AutoModelForCausalLM.from_pretrained(model_name)
        pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, device=device, max_length=max_length)
        return HuggingFacePipeline(pipeline=pipe)
    except Exception as e:
        logger.error(f"Failed to load LLM: {e}")
        return None

def process_pdf(file) -> Optional[List[Document]]:
    """Process the uploaded PDF file."""
    try:
        with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
            temp_file.write(file.getvalue())
            temp_file_path = temp_file.name

        loader = PyPDFLoader(file_path=temp_file_path)
        pages = loader.load()
        text_splitter = RecursiveCharacterTextSplitter(chunk_size=2000, chunk_overlap=100)
        documents = text_splitter.split_documents(pages)
        return documents
    except Exception as e:
        logger.error(f"Error processing PDF: {e}")
        return None

def create_vector_store(documents: List[Document], embeddings: HuggingFaceEmbeddings) -> Optional[FAISS]:
    """Create the vector store."""
    try:
        return FAISS.from_documents(documents, embeddings)
    except Exception as e:
        logger.error(f"Error creating vector store: {e}")
        return None

def summarize_report(documents: List[Document], llm: HuggingFacePipeline, max_length: int, summary_style: str) -> Optional[str]:
    """Summarize the report using the loaded model."""
    try:
        prompt_template = f"""
        Summarize the following text in a {summary_style} manner. Focus on the main points and key details:
        {{text}}
        Summary:
        """

        prompt = PromptTemplate(template=prompt_template, input_variables=["text"])
        chain = load_summarize_chain(llm, chain_type="stuff", prompt=prompt)
        summary = chain.run(documents, max_length=max_length)
        return summary

    except Exception as e:
        logger.error(f"Error summarizing report: {e}")
        return None

def main():
    st.title("Report Summarizer")

    model_option = st.sidebar.text_input("Enter model name", value=DEFAULT_MODEL)
    summary_style = st.sidebar.selectbox("Summary style", options=["clear and concise", "formal", "informal", "bullet points"])

    uploaded_file = st.sidebar.file_uploader("Upload your Report", type="pdf")

    llm = load_llm(model_option, 1024)  # Load the model with a default max_length
    if not llm:
        st.error(f"Failed to load the model {model_option}. Please try another model.")
        return

    embeddings = load_embeddings(EMBEDDING_MODEL)
    if not embeddings:
        st.error(f"Failed to load embeddings. Please try again later.")
        return

    if uploaded_file:
        with st.spinner("Processing PDF..."):
            documents = process_pdf(uploaded_file)

        if documents:
            with st.spinner("Creating vector store..."):
                db = create_vector_store(documents, embeddings)

            if db and st.button("Summarize"):
                # Calculate max_length based on input text
                input_length = sum([len(doc.page_content.split()) for doc in documents])
                max_length = int(input_length * MAX_LENGTH_FRACTION)

                # Reload the model with the calculated max_length
                llm = load_llm(model_option, max_length)

                with st.spinner(f"Generating summary using {model_option}..."):
                    summary = summarize_report(documents, llm, max_length, summary_style)

                    if summary:
                        st.subheader("Summary:")
                        st.write(summary)
                    else:
                        st.warning("Failed to generate summary. Please try again.")

if __name__ == "__main__":
    main()