How do I approach ML System Design interview questions?

ML System Design questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master ml system design interviews.

What difficulty level is this interview question?

This is a medium difficulty ML System Design question, commonly asked during Technical Screen rounds at Luma AI.

What role is this question designed for?

This question is commonly asked for Machine Learning Engineer candidates at Luma AI during technical interviews.

Debug MNIST denoiser training | Luma AI Interview Question

Quick Overview

Debug MNIST denoiser training evaluates ML product requirements, data/labeling, modeling, serving architecture, evaluation, monitoring, and trade-offs in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

Debug MNIST denoiser training

Debugging a Colab Denoising Network on MNIST

Goal: Make a Colab notebook that trains a denoising neural network on MNIST such that:

(a) the training loss steadily decreases, and
(b) the evaluation loss is close to the training loss.

You should identify and fix issues in three areas and describe how you validate each fix:

1) Data

Detect and correct a train/test distribution mismatch so the test set covers all digits 0–9.
Ensure the same normalization statistics (mean and std) computed from the training set are applied to both training and test data.

2) Model

Remove or replace an inappropriate final ReLU so the output range supports negative values typical of the denoised signal.

3) Optimization

Add the missing optimizer.zero_grad() and put backward() and optimizer.step() in the correct order.

For each of the three sections, describe:

Sanity checks and assertions you would add.
Minimal code changes and key code snippets.
Metrics you would track to validate the fix.

Constraints & Assumptions

Preserve the scope, facts, inputs, and requested outputs from the prompt above.
If the prompt leaves a detail unspecified, state a reasonable assumption before relying on it.
Keep the answer interview-ready: concise enough to present, but concrete enough to implement or evaluate.

Clarifying Questions to Ask

Clarify users, core use cases, read/write patterns, scale, latency, availability, and data retention.
State explicit assumptions before making sizing or architecture decisions.
Prioritize the functional path first, then address reliability, security, observability, and rollout.

What a Strong Answer Covers

A scoped requirements summary with concrete non-goals and success metrics.
ML-specific data, model, evaluation, serving, and monitoring choices.
Reasoned trade-offs among simple and scalable designs, including bottlenecks and failure modes.
A validation, monitoring, migration, and launch plan appropriate for the risk level.

Follow-up Questions

What breaks first at 10x traffic or data volume?
How would you degrade gracefully during dependency failures?
What metrics and alerts would prove the design is healthy after launch?

Quick Overview

3) Optimization

Add the missing optimizer.zero_grad() and put backward() and optimizer.step() in the correct order.

For each of the three sections, describe:

Sanity checks and assertions you would add.

Minimal code changes and key code snippets.

Metrics you would track to validate the fix.

Constraints & Assumptions

Preserve the scope, facts, inputs, and requested outputs from the prompt above.

If the prompt leaves a detail unspecified, state a reasonable assumption before relying on it.

Keep the answer interview-ready: concise enough to present, but concrete enough to implement or evaluate.

Clarifying Questions to Ask

Clarify users, core use cases, read/write patterns, scale, latency, availability, and data retention.

State explicit assumptions before making sizing or architecture decisions.

Prioritize the functional path first, then address reliability, security, observability, and rollout.

What a Strong Answer Covers

A scoped requirements summary with concrete non-goals and success metrics.

ML-specific data, model, evaluation, serving, and monitoring choices.

Reasoned trade-offs among simple and scalable designs, including bottlenecks and failure modes.

A validation, monitoring, migration, and launch plan appropriate for the risk level.

Follow-up Questions

What breaks first at 10x traffic or data volume?

How would you degrade gracefully during dependency failures?

What metrics and alerts would prove the design is healthy after launch?

Debug MNIST denoiser training

Quick Overview

Debug MNIST denoiser training

Debug MNIST denoiser training

Debugging a Colab Denoising Network on MNIST

1) Data

2) Model

3) Optimization

Constraints & Assumptions

Clarifying Questions to Ask

What a Strong Answer Covers

Follow-up Questions

Submit Your Answer to Earn 20XP

Debug MNIST denoiser training

Quick Overview

Debug MNIST denoiser training

Debug MNIST denoiser training

Debugging a Colab Denoising Network on MNIST

1) Data

2) Model

3) Optimization

Constraints & Assumptions

Clarifying Questions to Ask

What a Strong Answer Covers

Follow-up Questions

Submit Your Answer to Earn 20XP