xxuanwang-zz / Flask-app-with-Product-Type

Flask-app-with-Product-Type

This is a showcase of deploying a machine learning model to the web with Falsk on AWS EC2.

File Structure of the Flask App:

├── static  
│   ├── css  
│   │   ├── style.css  
├── templates  
│   ├── index.html  
├── images  
│   ├── flask_1.png  
|   ├── flask_2.png  
├── pt-app.py  
├── functions.py  
├── product_type_vectorizer.sav  
└── product_type_finalized_model.sav  
├── README.md
├── steps.txt 
└── requirements.txt

Project Structure

1. Model Deserialization

These two file are serialized from the NLP process within this repo. Here we'll use these files after de-serialized the pickled models in the form of python object.

• product_type_vectorizer.sav
• product_type_finalized_model.sav

2. Website template

• static/style.css

  • It contains the CSS styling of the main page.

• templates/index.html

  • It allows users enter the name and description of a makeup product, and show the prediction.

3. Flask Back-End API

• pt-app.py
  • Clean the input makeup details (text) with functions.py;
  • Predict the product type based on above mentioned models;
  • Display the output result on the GUI.

4. AWS EC2 Connection

• EC2 Instance

Other files

• steps.txt

  • It contains the commands to set up a virtual env, transfer files bwtween local machine and remote EC2 instance, the application url, and the test input of a makeup product.

• requirements.txt

  • Includes all the required python libraries/packages to run this app.

References:

• Deploy ML models on AWS

1. Deploying ML models To the Web with Flask on AWS EC2 Instance

2. 5 Different Ways to Deploy your Machine Learning Model with AWS

• Deploy ML models locally

3. Simple way to deploy machine learning models to cloud