Container registries for ML in MLOps - Mini Project: Build & Apply

Container Registries for ML

📖 Scenario: You are working on a machine learning project where you need to package your ML model into a container. To share and deploy this container easily, you will use a container registry. This project will guide you through creating a simple container image tag list, setting up a registry URL, filtering images by a tag prefix, and finally displaying the filtered images ready for deployment.

🎯 Goal: Build a small program that manages ML container image tags, filters them by a prefix, and prints the filtered list. This simulates preparing ML containers for deployment using a container registry.

📋 What You'll Learn

Create a dictionary of ML container images with exact tags and versions

Add a variable for the container registry URL

Filter the images to only those starting with a specific prefix

Print the filtered list of image tags

💡 Why This Matters

🌍 Real World

Container registries store and manage container images for ML models, making it easy to share and deploy models consistently.

💼 Career

Understanding container registries and filtering images is essential for ML engineers and DevOps professionals working on model deployment pipelines.

1

Create ML container images dictionary

Create a dictionary called ml_images with these exact entries: 'model_v1': 'ml-model:1.0', 'model_v2': 'ml-model:2.0', 'data_prep': 'data-prep:latest', 'model_experiment': 'ml-model-experiment:0.1'

MLOps

# Create the ml_images dictionary with exact keys and values
# Your code here

Need a hint?

Use curly braces {} to create a dictionary with the exact keys and values given.

2

Add container registry URL

Add a variable called registry_url and set it to the string 'registry.example.com/ml'

MLOps

ml_images = {
    'model_v1': 'ml-model:1.0',
    'model_v2': 'ml-model:2.0',
    'data_prep': 'data-prep:latest',
    'model_experiment': 'ml-model-experiment:0.1'
}
# Add the registry_url variable below
# Your code here

Need a hint?

Assign the exact string to the variable registry_url.

3

Filter images by prefix

Create a list called filtered_images that contains only the image tags from ml_images values starting with the prefix 'ml-model:'. Use a list comprehension.

MLOps

ml_images = {
    'model_v1': 'ml-model:1.0',
    'model_v2': 'ml-model:2.0',
    'data_prep': 'data-prep:latest',
    'model_experiment': 'ml-model-experiment:0.1'
}
registry_url = 'registry.example.com/ml'
# Create filtered_images list with tags starting with 'ml-model:'
# Your code here

Need a hint?

Use a list comprehension to check each tag in ml_images.values() and keep only those starting with 'ml-model:'.

4

Print filtered image tags

Print the filtered_images list to display the filtered ML container image tags.

MLOps

ml_images = {
    'model_v1': 'ml-model:1.0',
    'model_v2': 'ml-model:2.0',
    'data_prep': 'data-prep:latest',
    'model_experiment': 'ml-model-experiment:0.1'
}
registry_url = 'registry.example.com/ml'
filtered_images = [tag for tag in ml_images.values() if tag.startswith('ml-model:')]
# Print the filtered_images list below
# Your code here

Need a hint?

Use print(filtered_images) to show the filtered list.