Model stages (staging, production, archived) in MLOps - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Model Stages Purpose

Which statement best describes the purpose of the staging stage in a model lifecycle?

A It is where the model is tested in an environment similar to production before full deployment.

B It is the final stage where the model is actively used to make predictions for users.

C It is where old models are stored and no longer used for predictions.

D It is the initial phase where the model is being trained and developed.

💡 Hint

Think about the stage that acts as a 'practice run' before the model goes live.

The staging stage is used to test the model in an environment that mimics production. This helps catch issues before the model is fully deployed.

💻 Command Output

intermediate

Identifying Model Stage from CLI Output

You run a command to list models and their stages. What is the stage of the model named sales_forecast_v2?

MLOps

Model Name          Stage
sales_forecast_v1   archived
sales_forecast_v2   production
sales_forecast_v3   staging

A staging

B archived

C production

D training

💡 Hint

Look at the row with the model name sales_forecast_v2.

The output shows sales_forecast_v2 is in the production stage, meaning it is actively used for predictions.

🔀 Workflow

advanced

Correct Workflow for Promoting a Model to Production

Which sequence correctly describes the workflow to promote a model from staging to production?

A 1,2,3,4

B 2,1,4,3

C 1,3,2,4

D 3,1,2,4

💡 Hint

Think about testing before deploying and cleaning up old models after deployment.

The correct workflow is to test in staging, deploy to production, archive old models, then monitor performance.

❓ Troubleshoot

advanced

Troubleshooting Model Stage Mislabeling

You notice a model labeled as production is actually outdated and should not serve predictions. What is the best immediate action?

A Ignore the label and continue using the model.

B Delete the model files from storage immediately.

C Retrain the model and keep the stage as production.

D Change the model stage to <code>archived</code> to prevent it from being used.

💡 Hint

Think about how to stop the model from being used without losing data.

Changing the stage to archived prevents the model from being used while keeping it for records or rollback.

✅ Best Practice

expert

Best Practice for Managing Archived Models

What is the best practice for handling archived models in a production system?

A Use archived models as the default for new predictions.

B Keep archived models in a separate storage with restricted access for audit and rollback purposes.

C Move archived models back to staging for retraining without review.

D Delete archived models immediately to save storage space.

💡 Hint

Consider the importance of old models for audits and recovery.

Archived models should be stored safely with limited access to allow audits and rollback if needed.