Prototype Scikit-learn ML Model with Streamlit App

In this step-by-step tutorial, learn how to build, visualize, and deploy a Scikit-learn machine learning model using Streamlit for the UI, Docker for containerization, and Kubernetes (Minikube) for scalable deployment!

YouTube Link: https://youtu.be/Ska5_d63mLM?si=Xug9kxJoxheblrsu

Requirements

1. Python
2. Streamlit
3. Docker
4. DockerHub Account
5. Kubernetes Cluster (Minikube)

Install Required Packages

sudo apt update && sudo apt upgrade -y

sudo apt install python3-pip -y

Create Python Virtual Environment & Activate it

python -m venv venv
source venv/bin/activate

pip3 install streamlit scikit-learn pandas matplotlib seaborn reportlab

Run Streamlit App

streamlit run app.py

On your browser

http://localhost:8501

or

http://PublicIP:8501

Build and Run the Docker Image Using the Dockerfile

docker build -t streamlit-ml-app:latest .

docker run -p 8501:8501 streamlit-ml-app:latest

On your browser

http://localhost:8501

Push Streamlit Docker Image to DockerHub

Go to your DockerHub account and create a Personal Access Token (PAT) - This will be your password -

docker login -u iquantc

Tag Your Local Docker Image

docker tag streamlit-ml-app:latest iquantc/streamlit-ml-app:latest

Push the Image to DockerHub

docker push iquantc/streamlit-ml-app:latest

Deploy Streamlit App Docker Image from DockerHub to Kubernetes

Review manifest files

Create Minikube cluster

minikube start --driver=docker

Deploy the Kubernetes Manifest Files

kubectl apply -f deployment.yaml

kubectl apply -f service.yaml

Check the Resources Created

kubectl get pods

kubectl get svc

minikube ip

Open in Browser

http://<minikube-ip>:30001

Clean up

minikube stop

minikube delete --all