easyMultiple Choice
MLA-C01 Practice Question: A data scientist is training a deep learning…
A data scientist is training a deep learning model on SageMaker and notices that the training loss oscillates and does not converge. They want to debug this issue. Which SageMaker feature can they use to monitor and analyze the training process?
⚠ Common exam trap
AWS often tests the distinction between monitoring training metrics (Debugger) versus optimizing hyperparameters (Automatic Model Tuning) or profiling system resources (Profiler), leading candidates to confuse Debugger with tuning or profiling features.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
SageMaker Debugger
SageMaker Debugger is the correct feature because it provides real-time monitoring and analysis of training metrics, including loss values, gradients, and weights. It can automatically detect issues like oscillating or non-converging loss by setting rules (e.g., loss not decreasing) and emit alerts or capture tensors for later analysis, directly addressing the data scientist's need to debug training instability.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
SageMaker Profiler
Why it's wrong here
SageMaker Profiler reports hardware resource utilisation — GPU, CPU and memory bottlenecks — not loss curves or gradient statistics. It is tempting because profiling can reveal throughput problems, but diagnosing oscillating loss requires SageMaker Debugger's built-in tensor and loss monitoring rules.
- ✗
SageMaker Gradient Descent optimization
Why it's wrong here
Gradient descent is a training algorithm, not a SageMaker monitoring feature; no such SageMaker capability exists for inspecting convergence. It is tempting because oscillating loss often stems from learning-rate or optimiser settings, which you would tune manually, but the question asks which feature monitors and analyses the training process.
- ✓
SageMaker Debugger
Why this is correct
SageMaker Debugger captures training tensors and metrics in real time, letting the data scientist inspect loss curves and detect vanishing gradients, exploding gradients or poor learning rates. This directly addresses the oscillating, non-converging loss by exposing per-step training telemetry rather than only post-training evaluation.
- ✗
SageMaker Automatic Model Tuning
Why it's wrong here
Automatic Model Tuning runs hyperparameter search jobs to optimise objective metrics; it does not capture per-iteration loss curves, gradients or resource utilisation needed to diagnose oscillation. It is the right feature when the goal is improving accuracy through hyperparameter selection, not debugging convergence during training.
Go deeper
Related to this question
About these practice questions
One of 665 original MLA-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This MLA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLA-C01 exam.