Understanding the Workflow of Heroku Deployment

There are a set of steps to follow in order to deploy an app onto the Heroku Platform:

1. Make a web interface for our model:

• It is quite obvious that we should be having an app, for us it host it publicly :P.

• So to host a machine learning or deep learning model, the first step is to train a model.

• Then, we create a web interface, often using frameworks like Django or Flask, and make the app and web-app.

• We have already done all this in the project How to Deploy an Image Classification Model using Flask.

2. Create the requirements.txt file:

• A requirements.txt file is where we mention the details of all the packages - along with their corresponding versions - which are used in the project.

• This is a very important step in every real-case scenario because the app runs smoothly only if its dependencies are correctly installed.

• Similarly, Heroku relies on this requirements.txt file to install the dependencies and makes all the necessary arrangements to host the app.

3. Make the Procfile:

• The Procfile is like a configuration file to our app.

• It mentions to Heroku the entry-point for the app.

• Hence we shall define this file.

4. Commit the code on Git and push it to GitHub:

• GitHub is a version-control platform to host the source codes of our projects.

• To host our app on Heroku, we link our project directory - along with the Procfile and requirements.txt - which we push to GitHub.

5. Signup on Heroku Platform and Setting up Heroku::

• We need to create an account on Heroku.

• Once we log-in, we choose the option to link the GitHub repository to Heroku.

• Heroku then provides us the option to deploy(either automatically or manually). It installs the necessary dependencies as per the requirements.txt and sets all the configurations as per the Procfile.

• After Heroku does all such background work to host our app, it provides the link to the app.

• We could browse through that link and use our app.