Docker for ML workloads in MLOps - Mini Project: Build & Apply

Docker for ML Workloads

📖 Scenario: You are a data scientist who wants to share a machine learning model with your team. To make sure everyone can run the model easily, you decide to use Docker. Docker will package your model and its environment so it works the same on any computer.

🎯 Goal: Build a simple Docker setup that packages a Python script for a machine learning model. You will create the Python script, write a Dockerfile to set up the environment, build the Docker image, and run the container to see the model output.

📋 What You'll Learn

Create a Python script named model.py that prints a simple message simulating a model prediction.

Write a Dockerfile that uses a Python base image and copies model.py into the container.

Build a Docker image named ml-model from the Dockerfile.

Run a Docker container from the ml-model image and display the output.

💡 Why This Matters

🌍 Real World

Docker helps data scientists share ML models easily by packaging code and environment together. This avoids 'it works on my machine' problems.

💼 Career

Understanding Docker is essential for ML engineers and data scientists to deploy models reliably and collaborate with teams.

1

Create the Python script for the ML model

Create a file named model.py with a Python print statement that outputs exactly: "Model prediction: 42".

MLOps

# Create model.py with a print statement
# Your code here

Need a hint?

Use the print() function to display the message.

2

Write the Dockerfile to set up the environment

Create a Dockerfile that uses the Python 3.12 base image, copies model.py into the container, and sets the default command to run python model.py.

MLOps

# Write Dockerfile content below
# Your code here

Need a hint?

Use FROM to specify the base image, COPY to add files, WORKDIR to set the folder, and CMD to run the script.

3

Build the Docker image named ml-model

Run the Docker build command to create an image named ml-model using the current directory as context.

MLOps

# Run the docker build command below
# Your code here

Need a hint?

Use docker build -t ml-model . to build the image with the tag ml-model.

4

Run the Docker container and display the output

Run a Docker container from the ml-model image using the command docker run ml-model and display the output.

MLOps

# Run the docker container below
# Your code here

Need a hint?

Use docker run ml-model to start the container and see the printed message.