app.py

from pathlib import Path
from llama_index import ServiceContext, download_loader, GPTVectorStoreIndex, LLMPredictor
from llama_index import StorageContext, load_index_from_storage
from langchain import OpenAI
import gradio as gr
import os

# My personal OpenAI API key - do not misuse :P
os.environ["OPENAI_API_KEY"] = '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'

num_outputs = 512
llm_predictor = LLMPredictor(llm=OpenAI(temperature=0.4, model_name="text-davinci-003", max_tokens=num_outputs))
service_context = ServiceContext.from_defaults(llm_predictor=llm_predictor)

def construct_index():
    SimpleCSVReader = download_loader("SimpleCSVReader")
    loader = SimpleCSVReader()
    docs = loader.load_data(file=Path('./docs/articles.csv'))

    index = GPTVectorStoreIndex.from_documents(docs, service_context=service_context)

    index.storage_context.persist()

    return index

def chatbot(input_text):
    storage_context = StorageContext.from_defaults(persist_dir=Path("./storage/"))
    index = load_index_from_storage(storage_context, service_context=service_context)
    query_engine = index.as_query_engine(response_mode='compact')
    response = query_engine.query(input_text)
    return response.response

# uncomment on first run, when generating an index
index = construct_index()

iface = gr.Interface(fn=chatbot,
                     inputs=gr.inputs.Textbox(lines=7, label="Enter your text"),
                     outputs="text",
                     title="147 AI Chatbot")


iface.launch()

# with gr.Blocks() as demo:
#     chatbot = gr.Chatbot()
#     msg = gr.Textbox()
#     clear = gr.Button("Clear")

#     def respond(message, chat_history):
#         storage_context = StorageContext.from_defaults(persist_dir=Path("./storage/"))
#         index = load_index_from_storage(storage_context, service_context=service_context)
#         query_engine = index.as_query_engine(response_mode='compact')
#         response = query_engine.query(message)
#         chat_history.append((message, response))
#         return "", chat_history

#     msg.submit(respond, [msg, chatbot], [msg, chatbot])
#     clear.click(lambda: None, None, chatbot, queue=False)

# demo.launch()