FirstTry

Objective 3.2

PMLE

Training models

Objective 3.2 sits in Scaling prototypes into ML models, which carries 21% of the ML Engineer exam. The questions below are original, written from the official objective title above, and each explanation cites the Google Cloud page it rests on.

Objective title verbatim from the official objectives. Google Cloud exam page ↗

A worked example

Shown solved, with the whole explanation open: this is what every question here carries.

3-2Scaling prototypes into ML models

An engineer lays out four worker pools for a distributed job and needs the parameter servers in the right place. Which position holds them?

Position zero, alongside the primary replicaThat slot is reserved for the coordinating replica.
Position two, shared with Reduction ServerCorrect · your answerCorrect. That slot doubles for gradient aggregation.
Position one, with the worker replicasOrdinary workers occupy that position instead.
Position three, next to the evaluatorsEvaluation replicas sit at the end of the list.

Correct.

Checked against cloud.google.com, August 2026

Concept

In a cluster described positionally, the index carries the role. Getting a group of replicas into the right slot is not a naming choice; it is the only way the scheduler learns what that group is for.

Why B

The worker pool specifications map by position: the first is primary, chief, scheduler or master, the second is secondary replicas and workers, the third is parameter servers or Reduction Server, and the fourth is evaluators.

Source

Position in workerPoolSpecs[] Task performed in cluster First ( workerPoolSpecs[0] ) Primary, chief, scheduler, or "master" Second ( workerPoolSpecs[1] ) Secondary, replicas, workers Third ( workerPoolSpecs[2] ) Parameter servers, Reduction Server Fourth ( workerPoolSpecs[3] ) Evaluators…

Distributed training, Agent Platform documentation, checked August 2026
#gcp#distributed-training#cluster#configuration

Now you: objective 3.2 questions

No account needed. The explanation opens when you answer.

Sample question 1 of 3

3-2Scaling prototypes into ML models

You configure the first worker pool of a distributed Vertex AI training job and must set its replica count. What rule applies to that pool?

Sample question 2 of 3

3-2Scaling prototypes into ML models

You add Reduction Server to a multi-node GPU training job and must choose the machine type and node count for its worker pool. What applies to those nodes?

Sample question 3 of 3

3-2Scaling prototypes into ML models

A team wants Reduction Server on a training job whose workers all run on CPUs alone. What blocks that plan?

Full ML Engineer question bank coming

We’re writing the complete bank from the official objectives right now. Leave your email and we’ll tell you when it ships, nothing else, ever.

Read the sources

These are the official pages the questions above cite. Reading them is studying the objective from the primary source, which is what the explanations point you toward anyway.

More objectives in Scaling prototypes into ML models