Deploy a web app created with Streamlit to Heroku
I've created a web page that makes it easy to process images using Streamlit and Heroku, so I'll leave that introduction.
The app I made
Demo operation
Work log: Create an image processing app with streamlit. :: tomowarkar's tech blog — go 3 steps and go 2 steps
Run locally
pip install streamlit
streamlit run https://github.com/tomowarkar/stapp/blob/master/app.py
- Some functions such as face detection cannot be used.
What is Streamlit
Streamlit — The fastest way to build custom ML tools
Streamlit's open source app framework is the easiest way for data scientists and machine learning engineers to create beautiful, high-performance apps in hours. Everything in pure Python. It's all free. Streamlit — The fastest way to build custom ML tools (Google Japanese translation)
Goal of this article
--Deploy a web app made with Streamlit on Heroku.
Premise
-Heroku account and Heroku CLI need to be prepared
--The library is managed using pipennv, but of course you can use requirements.txt.
Create workspace, install streamlit
$ mkdir stapp
$ cd stapp
$ pipenv install --python 3
$ pipenv install streamlit
When using requirements.txt
$ mkdir stapp
$ cd stapp
$ pip insatll streamlit
$ echo streamlit==0.60.0 >requirements.txt
App creation
$ touch app.py
app.py
import streamlit as st
st.title("Hello Streamlit!!")
st.subheader("This is calculator.")
a = st.slider("a: ", 0, 10, 5, 1)
b = st.slider("b: ", 0, 10, 5, 1)
st.write(f"{a} x {b} = {a*b}")
That's it
Settings for Heroku [^ 1]
$ echo "web: streamlit run --server.enableCORS false --server.port \$PORT app.py" >Procfile
Git repository initialization
$ git init
$ git add .
$ git commit -m "first commit"
Creating a Heroku app
$ heroku create
$ git push heroku master
$ heroku open
Was the website displayed safely?
Next step
Tutorial: Create a data explorer app — Streamlit 0.61.0 documentation
Streamlit's official tutorial is just right and interesting!
in conclusion
What did you think. You can publish your web app in less than 10 minutes at the earliest.
Although streamlit is not very customizable, you can easily create a web application with a table created with DataFrame of pandas, a graph plotted with matplotlib, and the result of image processing with ʻOpenCV` in just a few lines. You can drop it.
It is very nice that it is completed only with python and you can easily output the deliverables.
Recommended Posts