Google Professional Machine Learning Engineer
Get started today
Ultimate access to all questions.
You are responsible for deploying a new scikit-learn classification model into a production environment for a company application. The system must be reliable and capable of handling high traffic loads, processing millions of requests per second from 8 am to 7 pm every day. Given these requirements, you must also focus on minimizing overall deployment costs while ensuring the system's performance and uptime. What should you do?
You are responsible for deploying a new scikit-learn classification model into a production environment for a company application. The system must be reliable and capable of handling high traffic loads, processing millions of requests per second from 8 am to 7 pm every day. Given these requirements, you must also focus on minimizing overall deployment costs while ensuring the system's performance and uptime. What should you do?
Explanation:
The correct answer is B. You don't need a GPU for a scikit-learn model because scikit-learn doesn't support GPU acceleration. Setting the max replica count to 100 allows the system to scale and handle the high volume of requests efficiently. While it might be more expensive, it ensures that your system can handle the expected traffic load. Option A with a max replica count of 1 would not be able to handle millions of requests per second. Options C and D are incorrect because they unnecessarily include GPU, which is not needed for scikit-learn models.