app.py

import os
from dotenv import load_dotenv
import gradio as gr
from llama_index.core import StorageContext, load_index_from_storage, VectorStoreIndex, SimpleDirectoryReader, ChatPromptTemplate, Settings
from llama_index.llms.huggingface import HuggingFaceInferenceAPI
from llama_index.embeddings.huggingface import HuggingFaceEmbedding
from sentence_transformers import SentenceTransformer

load_dotenv()

# Configure the Llama index settings
Settings.llm = HuggingFaceInferenceAPI(
    model_name="google/gemma-1.1-7b-it",
    tokenizer_name="google/gemma-1.1-7b-it",
    context_window=3000,
    token=os.getenv("HF_TOKEN"),
    max_new_tokens=512,
    generate_kwargs={"temperature": 0.1},
)
Settings.embed_model = HuggingFaceEmbedding(
    model_name="BAAI/bge-small-en-v1.5"
)

# Define the directory for persistent storage and data
PERSIST_DIR = "db"
PDF_DIRECTORY = 'data'  # Changed to the directory containing PDFs

# Ensure PDF directory exists
os.makedirs(PDF_DIRECTORY, exist_ok=True)
os.makedirs(PERSIST_DIR, exist_ok=True)


def data_ingestion_from_directory():
    # Use SimpleDirectoryReader on the directory containing the PDF files
    documents = SimpleDirectoryReader(PDF_DIRECTORY).load_data()
    storage_context = StorageContext.from_defaults()
    index = VectorStoreIndex.from_documents(documents)
    index.storage_context.persist(persist_dir=PERSIST_DIR)


def handle_query(query, history):
    chat_text_qa_msgs = [
        (
            "user",
            """
            You are a RedfernsTech chatbot whose aim is to provide better service to the user, utilizing provided context to deliver answers.
            and collect some basic information first like name, email, company name.
            {context_str}
            Question:
            {query_str}
            """
        )
    ]
    text_qa_template = ChatPromptTemplate.from_messages(chat_text_qa_msgs)

    # Load index from storage
    storage_context = StorageContext.from_defaults(persist_dir=PERSIST_DIR)
    index = load_index_from_storage(storage_context)

    query_engine = index.as_query_engine(text_qa_template=text_qa_template)
    answer = query_engine.query(query)

    if hasattr(answer, 'response'):
        response = answer.response
    elif isinstance(answer, dict) and 'response' in answer:
        response = answer['response']
    else:
        response = "Sorry, I couldn't find an answer."

    history.append((query, response))
    return response, history


# Example usage
print("Processing PDF ingestion from directory:", PDF_DIRECTORY)
data_ingestion_from_directory()

# Example query
query = "How do I use the RedfernsTech Q&A assistant?"
print("Query:", query)
response = handle_query(query, [])
print("Answer:", response)

# Create the Gradio chatbot interface with history
chatbot = gr.ChatInterface(
    fn=handle_query,
    title="RedfernsTech Q&A Chatbot",
    description="Ask me anything about the uploaded documents.",
    cache_examples=True,  # Enable history caching
)

chatbot.launch()