From the course: Docker for Data Engineers

Unlock this course with a free trial

Join today to access over 24,900 courses taught by industry experts.

Setting up the Dockerfile for model prediction

Setting up the Dockerfile for model prediction - Docker Tutorial

From the course: Docker for Data Engineers

Start my 1-month free trial Buy for my team

Setting up the Dockerfile for model prediction

“

We have our classification model serialized out to a pickle file. Let me just copy that model over to our docker-demos folder. This is the folder that is our current working directory for building our Docker image to run a containerized application for predictions. Now, here I have all of the contents of my docker-demos folder open. Let's look at the files that we have that make up our prediction application. Take a look at the right side of the screen. Notice that I have the app.py Python file. This is my actual Flask Python application that we'll use to perform predictions behind an endpoint. We'll look at the contents of this app.py in just a bit. We have the churn_pred_model.pkl file, that is our serialized trained model. We have the Dockerfile, which specifies the instructions to build our image. We have the gunicorn_config.py file. Now, we haven't really worked with Gunicorn before, but this is what we'll use to run a production-grade web server hosting our Flask application…

Contents