Skip to main content

Building Apps with StreamLit

Overview​

Streamlit provides a quick and easy way for you to create an app for your AI model. The tool can automatically generate a website that incorporate all kinds of inputs and display widgets. You can consult the full documentation on Streamlit website.

In this section, we will walk through a simple demonstration of an image classification app built with Streamlit and PyTorch.

Install Streamlit​

Make sure you activate your warp environment with PyTorch installed. Then import the necessary packages for creating a simple neural network.

conda activate warp
conda install streamlit

Verify that you can successfully launch streamlit app. The following command should open up a web browser pointing to your app.

streamlit hello

A Simple App​

Streamlit can 'magically' display your data on the web page. For example, you can turn the following simple python script into web app.

demo.py
import streamlit as st
import torch
import pandas as pd

# Draw a title and some text to the app:
'''
# This is the document title

This is some _markdown_.
'''

import pandas as pd
df = pd.DataFrame({'col1': [1,2,3],'col2': [4,5,6]})
df  # 👈 Draw the dataframe

x = torch.tensor([[1, 2, 3], [4, 5, 6]])
tf = pd.DataFrame(x)
tf # 👈 Draw the tensor as dataframe

x = 10
'x', x  # 👈 Draw the string 'x' and then the value of x

# Also works with most supported chart types
import matplotlib.pyplot as plt
import numpy as np

arr = np.random.normal(1, 1, size=100)
fig, ax = plt.subplots()
ax.hist(arr, bins=20)

fig  # 👈 Draw a Matplotlib chart

You can then launch the sample app.

streamlit run demo.py

Acknowledgement : The content of this document has been adapted from these original websites.