app.py

from llama_index.core import (
    Settings,
    load_index_from_storage,
    StorageContext,
)
from llama_index.core.node_parser import SentenceSplitter
from llama_index.llms.openai import OpenAI 
import gradio as gr
import os

"""
Code adopted from Beebom article: "How to Train an AI Chatbot With Custom Knowledge Base Using ChatGPT API" by Arjun Sha
https://beebom.com/how-train-ai-chatbot-custom-knowledge-base-chatgpt-api/
Updated Feb 22, 2025 to use updated OpenAI API and Llama Index library
"""

max_input_size = 4096
num_outputs = 512
chunk_size_limit = 600
chunk_overlap = int(chunk_size_limit * 0.1) 

llm = OpenAI(
    model="gpt-3.5-turbo",
    temperature=0.5,
    max_tokens=num_outputs
)

Settings.llm = llm
Settings.node_parser = SentenceSplitter(
    chunk_size=chunk_size_limit,
    chunk_overlap=chunk_overlap
)
Settings.context_window = max_input_size
Settings.num_output = num_outputs

def retrieve_index(index_path):
    storage_context = StorageContext.from_defaults(persist_dir=index_path)
    index = load_index_from_storage(storage_context)
    return index

def chatbot(input_text):
    response = QE.query(input_text)
    response_stream = ""
    for r in response.response_gen:
        response_stream += r
        yield response_stream

if __name__ == "__main__":
    iface = gr.Interface(fn=chatbot,
                         inputs=gr.components.Textbox(lines=7, label="Enter your text"),
                         outputs="text",
                         title="AI Chatbot for the Doing What Works Library")

    index = retrieve_index("dww_vectors")
    QE = index.as_query_engine(streaming=True)
    iface.launch(share=False)