Deploying Machine Learning Models with Flask

Created by Vishal Verma in Articles 19 Feb 2024

Today, we're going to talk about making a super cool web app. Imagine a web app that can guess things using machine learning – how awesome is that? Don't worry if it sounds a bit tricky; we'll take it step by step, and you'll have a fantastic web app by the end.

What You Need

To get started, make sure you have two things:

1. Your computer with a special program called Python (version 3.6 or newer).
2. A little bit of knowledge about machine learning (the computer guessing things) and a tiny bit about Flask (a tool for making web apps).

Step 1: Install Some Stuff

Open a special window on your computer – we call it terminal. Then, type in some magic words:

python -m venv venv

Activate this magic space:

If you're using Windows, type venv\Scripts\activate.

If you're using MacOS/Linux, type source venv/bin/activate.

Now, let's add more magic:

pip install Flask scikit-learn

Step 2: Make Your Web Home

Imagine your web app having a home. Inside this home, there's a special file called app.py. This file is like the heart of your web app:

# app.py

from flask import Flask, render_template



app = Flask(__name__)



@app.route('/')

def home():

    return render_template('index.html')



if __name__ == '__main__':

    app.run(debug=True)

Step 3: Organize Your Stuff

Think of your web app like a tidy room. Make a folder called templates inside your web app's home. In this folder, make an HTML drawing called index.html. This drawing will be how your web app looks:

<html lang="en">

<head>

    <meta charset="UTF-8">

    <meta name="viewport" content="width=device-width, initial-scale=1.0">

    <title>Machine Learning Model Deploymenttitle>

<body>

    <h1>Welcome to our Machine Learning Deployment Apph1>

Step 4: Show Your Magic Model

Make another space in your web app's home called model. In this space, put a magic drawing (let's call it model.pkl). Now, tell your web app's heart about this magic:

# app.py

from flask import Flask, render_template

import joblib



app = Flask(__name__)



# Load the magic drawing

model = joblib.load('model/model.pkl')



@app.route('/')

def home():

    # Add a bit of magic to make guesses using the drawing

    # For example: prediction = model.predict([[input_data]])

    return render_template('index.html')

Step 5: See the Future (Prediction Place)

Tell your web app's heart to have a special place for guesses:

# app.py

from flask import Flask, render_template, request, jsonify

import joblib



app = Flask(__name__)



# Load the magic drawing

model = joblib.load('model/model.pkl')



@app.route('/')

def home():

    return render_template('index.html')



@app.route('/predict', methods=['POST'])

def predict():

    # Get special info from the request

    data = request.get_json(force=True)



    # Do magic guesses using the drawing

    # For example: prediction = model.predict([[input_data]])



    # Show the guess in a special way

    return jsonify({'prediction': 'your_prediction'})



if __name__ == '__main__':

    app.run(debug=True)

Flask Documentation

To learn more about Flask, you can explore the official documentation here.

Step 6: Try Your Web App

Start your web app:

python app.py

Go to http://127.0.0.1:5000/ on your computer to see your web app's first page.

In the next article, we will create a real-application that will predict the output based on user input. We will discuss the following in our next post:

1. Train a Machine Learning Model.
2. Create a Prediction Function.
3. Build an HTML-based form to collect user input.
4. Create functions to handle the user input, generate response, and display it in browser.

A glimpse of the application that we will be covering in our next article:

Application:

Result:

Conclusion

Congratulations! You've successfully created a simple Flask web application that can serve machine learning predictions. This tutorial provides a basic structure, and you can expand upon it by incorporating more complex machine learning models, enhancing the user interface, and deploying it to production servers. Flask's flexibility and simplicity make it an excellent choice for deploying machine learning models and building interactive applications.