QuiztudyPREMIUM
🌍 All Study Guides📊 Dashboard📰 Blog💡 About
AWS Certified AI Practitioner • STUDY MODE

Practice Quiz

QUESTION 1 OF 5

A machine learning developer is configuring an Amazon SageMaker batch transform job. They initialize multiple compute instances to process a single large input file. However, during execution, they notice that only one compute instance is active while the rest remain idle. They need to partition the input file into mini-batches to resolve this. Which configuration or behavior explains this scenario and provides the correct solution?

A
Since only a single input file is provided, only one compute instance processes it while the others remain idle. To split the file into mini-batches, the developer should set the SplitType parameter value to Line.Correct Answer
B
Batch transform requires the developer to set MaxPayloadInMB to over 100 MB to automatically split a single file across multiple instances.
C
SageMaker batch transform automatically splits single files by default, but only if the MaxConcurrentTransforms parameter is set to greater than 100 MB.
D
The developer must rename the output file to have a .out extension prior to the job execution to allow multiple instances to write to S3 concurrently.
Explanation:

According to SageMaker batch transform facts, initializing multiple compute instances with only one input file results in only one active instance while the rest are idle. To partition the file into mini-batches and utilize the instances, SplitType must be set to Line. The MaxPayloadInMB cannot exceed 100 MB, and output filenames are automatically appended with .out upon successful processing.

QUESTION 2 OF 5

An engineering team is configuring an Amazon SageMaker batch transform job. They want to set the optional MaxConcurrentTransforms parameter and the MaxPayloadInMB parameter to optimize throughput, and they need to know what the resulting S3 output file name will be after successful execution on an input file named dataset_records.csv. Which configuration and output naming behavior is correct?

A
The MaxPayloadInMB parameter must not exceed 100 MB, and the product of MaxConcurrentTransforms and MaxPayloadInMB must not exceed 100 MB. The successful output file will be named dataset_records.csv.out.Correct Answer
B
The MaxPayloadInMB parameter must not exceed 100 MB, and the sum of MaxConcurrentTransforms and MaxPayloadInMB must not exceed 100 MB. The output file will be named dataset_records.out.csv.
C
The MaxPayloadInMB parameter can be set up to 1000 MB as long as MaxConcurrentTransforms is set to 1. The successful output file will be named dataset_records.csv.out.
D
The MaxPayloadInMB parameter must not exceed 100 MB, and the product of MaxConcurrentTransforms and MaxPayloadInMB must be exactly 100 MB. The output file will be renamed dataset_records_out.csv.
Explanation:

In SageMaker batch transform, the MaxPayloadInMB parameter must not exceed 100 MB. If MaxConcurrentTransforms is set, the product of the two parameters must not exceed 100 MB. On successful execution, the output file retains the same name as the input file, with .out appended.

QUESTION 3 OF 5

An AI practitioner is evaluating options for processing dataset inference in Amazon SageMaker. The requirements are to preprocess datasets to remove noise or bias, get inferences from large datasets without deploying a persistent endpoint, and associate input records with inferences to help interpret results. Which SageMaker feature should be selected to meet these requirements?

A
Amazon SageMaker batch transformCorrect Answer
B
A persistent SageMaker real-time endpoint
C
A manual SageMaker model training job with persistent hosting
D
SageMaker live stream preprocessing
Explanation:

According to the facts, Amazon SageMaker batch transform is the correct feature to use when you need to preprocess datasets to remove noise or bias, get inferences from large datasets, run inference without a persistent endpoint, and associate input records with inferences to help with results interpretation.

QUESTION 4 OF 5

An ML specialist is running an Amazon SageMaker AI batch transform job to get inferences from a large dataset. The specialist configures the job to initialize three compute instances, but places only a single large CSV input file in the Amazon S3 input location. During execution, they notice that only one compute instance is actively processing the file while the other two remain idle. Why are the other two instances idle?

A
SageMaker AI batch transform partitions S3 input objects by key and maps them to instances, meaning only one instance can process a single file.Correct Answer
B
To distribute a single file, the SplitType parameter must be configured to Line, which triggers multi-instance file partitioning automatically.
C
The product of MaxConcurrentTransforms and MaxPayloadInMB has exceeded 100 MB, forcing SageMaker AI to default to a single instance.
D
Separate transform jobs with unique model names must be configured to enable multi-instance parallel processing of a single dataset.
Explanation:

When a batch transform job starts, SageMaker AI partitions the Amazon S3 objects in the input by key and maps those S3 objects to instances. If multiple compute instances are initialized but only a single input file exists, only one instance will process the file, leaving the other instances idle.

QUESTION 5 OF 5

When designing an Amazon SageMaker AI batch transform job, which constraint must be satisfied regarding the relationship between the optional MaxConcurrentTransforms parameter and the MaxPayloadInMB parameter?

A
The product of MaxConcurrentTransforms and MaxPayloadInMB must not exceed 100 MB.Correct Answer
B
The product of MaxConcurrentTransforms and MaxPayloadInMB must be configured to exactly 100 MB.
C
The SplitType parameter must be configured to Line if the product of these two parameters exceeds 100 MB.
D
Separate transform jobs must be created for each variant if the product of these two parameters exceeds 100 MB.
Explanation:

In SageMaker batch transform, mini-batch size is controlled using BatchStrategy and MaxPayloadInMB. MaxPayloadInMB cannot exceed 100 MB, and if MaxConcurrentTransforms is set, the product of MaxConcurrentTransforms and MaxPayloadInMB must not exceed 100 MB.

Ready to test your recall?

A machine learning developer is configuring an Amazon SageMaker batch transform job. They initialize multiple compute instances to process a single large input file. However, during execution, they notice that only one compute instance is active while the rest remain idle. They need to partition the input file into mini-batches to resolve this. Which configuration or behavior explains this scenario and provides the correct solution?

A
Since only a single input file is provided, only one compute instance processes it while the others remain idle. To split the file into mini-batches, the developer should set the SplitType parameter value to Line.
B
Batch transform requires the developer to set MaxPayloadInMB to over 100 MB to automatically split a single file across multiple instances.
C
SageMaker batch transform automatically splits single files by default, but only if the MaxConcurrentTransforms parameter is set to greater than 100 MB.
D
The developer must rename the output file to have a .out extension prior to the job execution to allow multiple instances to write to S3 concurrently.

How confident are you in this answer?