Answer-first summary for fast verification
Answer: DataflowPythonJobOp, WaitGcpResourcesOp, and CustomTrainingJobOp
The correct answer is B. The DataflowPythonJobOp component is well-suited for preprocessing tasks such as resizing images, converting them to grayscale, and extracting features, especially when these tasks need to be performed in parallel. The WaitGcpResourcesOp component ensures that the Dataflow job completes before moving to the next step. Finally, the CustomTrainingJobOp allows for custom training of the model once preprocessing is done. Other options are not suitable: A involves Spark which is not mentioned in the problem; C involves more manual orchestration and potentially less efficient for image preprocessing; D is oriented towards AutoML tasks and not suitable for custom preprocessing and training.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
As a machine learning engineer, you are developing a custom image classification model that will be implemented using Vertex AI Pipelines to manage the end-to-end processing and training workflow. Your dataset consists of images requiring several preprocessing steps before being suitable for training. These steps include resizing the images, converting them to grayscale, and extracting specific features. You have developed Python functions to handle these preprocessing tasks. Considering these requirements and the components available in Vertex AI Pipelines, which components should you use in your pipeline?
A
DataprocSparkBatchOp and CustomTrainingJobOp
B
DataflowPythonJobOp, WaitGcpResourcesOp, and CustomTrainingJobOp
C
dsl.ParallelFor, dsl.component, and CustomTrainingJobOp
D
ImageDatasetImportDataOp, dsl.component, and AutoMLImageTrainingJobRunOp