Skip to main content
Artificial Intelligence ⭐ Premium

Databricks Certified Machine Learning Associate - 340 Questions

By Webmaster Certland English ❤️ 0 likes

Practice exam for the Databricks Certified Machine Learning Associate certification. Covers Databricks ML platform, AutoML, MLflow tracking, Feature Store, model development with scikit-learn and PyTorch, and model deployment with Databricks Model Serving.

⭐ Premium Updated Mar 2026

Unlock all 340 Databricks Certified Machine Learning Associate - 340 Questions questions

Full simulation · Detailed explanations · Unlimited attempts

  • 340 questions — ~5 full-length simulations
  • Detailed explanations — why each answer is right or wrong
  • Unlimited attempts — retake as many times as needed
  • Smart Practice + Focus Mode + no ads
340
Questions
All certifications
from $4.90/mo

Sample Questions — Databricks Certified Machine Learning Associate - 340 Questions

5 free sample questions from this practice exam. Correct answers are highlighted.

1. A data scientist is setting up a new Databricks cluster for training scikit-learn and XGBoost models. They want to avoid manually installing these libraries. Which cluster configuration should they choose?

A Select a Databricks Runtime ML version (e.g., 14.3 ML) ✓ Correct
B Select a standard Databricks Runtime version (e.g., 14.3)
C Select a Databricks Runtime with Photon acceleration
D Select a single-node cluster with the standard runtime

2. A data scientist is using MLflow to track experiments. After a training run completes, the team wants to understand exactly what information MLflow stores automatically in the tracking server. Which of the following is NOT stored by the MLflow tracking server during a run?

A Hyperparameter values logged with mlflow.log_param()
B Evaluation metrics logged with mlflow.log_metric()
C Model files logged with mlflow.sklearn.log_model()
D The full training dataset used during model training ✓ Correct

3. Which of the following libraries is pre-installed in Databricks Runtime ML but NOT in the standard Databricks Runtime?

A pandas
B PySpark
C PyTorch ✓ Correct
D NumPy

4. A machine learning engineer needs to track multiple evaluation metrics over training epochs in MLflow. Which approach correctly logs a metric at each step?

A mlflow.log_param('loss', loss_value, step=epoch)
B mlflow.log_metric('loss', loss_value, step=epoch) ✓ Correct
C mlflow.log_artifact('loss', loss_value, step=epoch)
D mlflow.set_tag('loss', loss_value)

5. A machine learning engineer needs to fine-tune a large language model using GPU-accelerated training on Databricks. Which action is required to enable GPU usage on a cluster?

A Install CUDA manually via an init script before starting the cluster
B Select a GPU-enabled instance type when creating or editing the cluster ✓ Correct
C Enable Photon acceleration in the cluster configuration
D Use the standard Databricks Runtime and install GPU drivers as a cluster library

Want to test yourself for real?

Create a free account and run our exam simulation engine.

Free No credit card
  • Simulation engine
  • Up to 10 questions per attempt
  • Score & basic stats
Create free account Already have an account? Sign in
Best
Premium Premium
  • All 340 questions
  • Detailed explanations
  • Smart Practice + Focus Mode
⭐ Get Premium

Related Exams

Discussion

No comments yet. Be the first to start the discussion!

Sign in to join the discussion.