app.py

from dotenv import load_dotenv
import gradio as gr
import os
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
import firebase_admin
from firebase_admin import db, credentials
import datetime
import uuid
import random

def select_random_name():
    names = ['Clara', 'Lily']
    return random.choice(names)

# Example usage
# Load environment variables
load_dotenv()
# authenticate to firebase
cred = credentials.Certificate("redfernstech-fd8fe-firebase-adminsdk-g9vcn-0537b4efd6.json")
firebase_admin.initialize_app(cred, {"databaseURL": "https://redfernstech-fd8fe-default-rtdb.firebaseio.com/"})
# Configure the Llama index settings
Settings.llm = HuggingFaceInferenceAPI(
    model_name="meta-llama/Meta-Llama-3-8B-Instruct",
    tokenizer_name="meta-llama/Meta-Llama-3-8B-Instruct",
    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 directories exist
os.makedirs(PDF_DIRECTORY, exist_ok=True)
os.makedirs(PERSIST_DIR, exist_ok=True)

# Variable to store current chat conversation
current_chat_history = []
kkk=select_random_name()
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):
    chat_text_qa_msgs = [
        (
            "user",
            """
            As {kkk}, your goal is to offer top-tier service and information about RedFerns Tech company.
            Provide concise answers based on the conversation flow. Ultimately, aim to attract users to connect with our services.
            Summarize responses effectively in 20-60 words without unnecessary repetition.
            {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)

    # Use chat history to enhance response
    context_str = ""
    for past_query, response in reversed(current_chat_history):
        if past_query.strip():
            context_str += f"User asked: '{past_query}'\nBot answered: '{response}'\n"

    query_engine = index.as_query_engine(text_qa_template=text_qa_template, context_str=context_str)
    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."

    # Update current chat history
    current_chat_history.append((query, response))

    return response

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

# Define the function to handle predictions
"""def predict(message,history):
    response = handle_query(message)
    return response"""

def predict(message, history):
    logo_html = '''
    <div class="circle-logo">
      <img src="https://rb.gy/8r06eg" alt="FernAi">
    </div>
    '''
    response = handle_query(message)
    response_with_logo = f'<div class="response-with-logo">{logo_html}<div class="response-text">{response}</div></div>'
    return response_with_logo
def save_chat_message(session_id, message_data):
    ref = db.reference(f'/chat_history/{session_id}')  # Use the session ID to save chat data
    ref.push().set(message_data)

# Define your Gradio chat interface function (replace with your actual logic)
def chat_interface(message, history):
    try:
        # Generate a unique session ID for this chat session
        session_id = str(uuid.uuid4())

        # Process the user message and generate a response (your chatbot logic)
        response = handle_query(message)

        # Capture the message data
        message_data = {
            "sender": "user",
            "message": message,
            "response": response,
            "timestamp": datetime.datetime.now().isoformat()  # Use a library like datetime
        }

        # Call the save function to store in Firebase with the generated session ID
        save_chat_message(session_id, message_data)

        # Return the bot response
        return response
    except Exception as e:
        return str(e)

# Custom CSS for styling
css = '''
  .circle-logo {
  display: inline-block;
  width: 40px;
  height: 40px;
  border-radius: 50%;
  overflow: hidden;
  margin-right: 10px;
  vertical-align: middle;
}
.circle-logo img {
  width: 100%;
  height: 100%;
  object-fit: cover;
}
.response-with-logo {
  display: flex;
  align-items: center;
  margin-bottom: 10px;
}
footer {
    display: none !important;
    background-color: #F8D7DA;
  }
label.svelte-1b6s6s {display: none}
'''

gr.ChatInterface(chat_interface,
                 css=css,
                 description=kkk,
                 clear_btn=None, undo_btn=None, retry_btn=None,
                 ).launch()