decodingdatascience's picture
Create app.py
09f7248 verified
# app.py
import os
import logging
import gradio as gr
from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, StorageContext
from llama_index.vector_stores.pinecone import PineconeVectorStore
from pinecone import Pinecone, ServerlessSpec
# Logging setup
logging.basicConfig(level=logging.INFO)
api_key = os.environ["PINECONE_API_KEY"]
# Initialize Pinecone
pc = Pinecone(api_key=api_key)
index_name = "quickstart"
dimension = 1536
# Delete index if exists (optional)
if index_name in [idx['name'] for idx in pc.list_indexes()]:
pc.delete_index(index_name)
# Create new index
pc.create_index(
name=index_name,
dimension=dimension,
metric="euclidean",
spec=ServerlessSpec(cloud="aws", region="us-east-1"),
)
pinecone_index = pc.Index(index_name)
# Download data if not exists
os.makedirs("data/paul_graham", exist_ok=True)
file_path = "data/paul_graham/paul_graham_essay.txt"
if not os.path.exists(file_path):
import urllib.request
urllib.request.urlretrieve(
"https://raw.githubusercontent.com/run-llama/llama_index/main/docs/docs/examples/data/paul_graham/paul_graham_essay.txt",
file_path
)
# Load documents
documents = SimpleDirectoryReader("data/paul_graham/").load_data()
# Build vector index
vector_store = PineconeVectorStore(pinecone_index=pinecone_index)
storage_context = StorageContext.from_defaults(vector_store=vector_store)
index = VectorStoreIndex.from_documents(documents, storage_context=storage_context)
query_engine = index.as_query_engine()
# Gradio UI function
def query_doc(prompt):
try:
response = query_engine.query(prompt)
return str(response)
except Exception as e:
return f"Error: {str(e)}"
# Launch Gradio app
gr.Interface(
fn=query_doc,
inputs=gr.Textbox(label="Ask a question about the document"),
outputs=gr.Textbox(label="Answer"),
title="Paul Graham Document QA (LlamaIndex + Pinecone)",
description="Ask questions based on the indexed Paul Graham essay. Powered by LlamaIndex & Pinecone."
).launch()