Answer-first summary for fast verification
Answer: Utilize the AI Platform's custom containers feature to accept training jobs across any framework, including custom libraries, offering flexibility and reducing management overhead., Configure Kubeflow on Google Kubernetes Engine to handle training jobs via TFJob, ensuring compatibility with TensorFlow-based frameworks but potentially limiting support for others.
The AI Platform custom containers feature is the optimal solution because it supports training jobs from any framework, including custom libraries, thus simplifying the management of diverse frameworks. This approach not only adheres to cost constraints by reducing the need for multiple VM images but also ensures scalability for future framework additions. The combination with Kubeflow (option E) is also correct for teams heavily invested in TensorFlow, offering optimized performance for TensorFlow-based jobs while maintaining flexibility for other frameworks.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
As a manager of a data science team utilizing a complex cloud-based backend for training jobs, you're facing management challenges due to the team's use of diverse frameworks like Keras, PyTorch, Theano, Scikit-learn, and custom libraries. The solution must not only streamline the process but also adhere to cost constraints and ensure scalability for future framework additions. Which managed service would best meet these requirements? Choose the best option.
A
Configure Kubeflow on Google Kubernetes Engine to handle training jobs via TFJob, ensuring compatibility with TensorFlow-based frameworks but potentially limiting support for others.
B
Utilize the AI Platform's custom containers feature to accept training jobs across any framework, including custom libraries, offering flexibility and reducing management overhead.
C
Establish a Slurm workload manager for scheduling and executing jobs on your cloud infrastructure, which requires significant setup and maintenance but offers fine-grained control.
D
Develop a library of VM images on Compute Engine, making them available in a centralized repository, which could increase costs and complexity due to the need for multiple images for different frameworks.
E
Implement a combination of AI Platform's custom containers for most frameworks and Kubeflow for TensorFlow-based jobs, ensuring broad compatibility and optimized performance where possible. Choose two correct options.