Deal of The Day! Hurry Up, Grab the Special Discount - Save 25% - Ends In 00:00:00 Coupon code: SAVE25
Welcome to Pass4Success

- Free Preparation Discussions
Mail Us support@pass4success.com
Location US
Mob_nav_barsMENU

Google Professional Machine Learning Engineer Exam - Topic 10 Question 58 Discussion

You have been asked to productionize a proof-of-concept ML model built using Keras. The model was trained in a Jupyter notebook on a data scientist's local machine. The notebook contains a cell that performs data validation and a cell that performs model analysis. You need to orchestrate the steps contained in the notebook and automate the execution of these steps for weekly retraining. You expect much more training data in the future. You want your solution to take advantage of managed services while minimizing cost. What should you do?
D) Extract the steps contained in the Jupyter notebook as Python scripts, wrap each script in an Apache Airflow BashOperator, and run the resulting directed acyclic graph (DAG) in Cloud Composer.
A) Move the Jupyter notebook to a Notebooks instance on the largest N2 machine type, and schedule the execution of the steps in the Notebooks instance using Cloud Scheduler.
B) Write the code as a TensorFlow Extended (TFX) pipeline orchestrated with Vertex AI Pipelines. Use standard TFX components for data validation and model analysis, and use Vertex AI Pipelines for model retraining.
C) Rewrite the steps in the Jupyter notebook as an Apache Spark job, and schedule the execution of the job on ephemeral Dataproc clusters using Cloud Scheduler.

Google Professional Machine Learning Engineer Exam - Topic 10 Question 58 Discussion

Actual exam question for Google's Professional Machine Learning Engineer exam
Question #: 58
Topic #: 10
[All Professional Machine Learning Engineer Questions]

You have been asked to productionize a proof-of-concept ML model built using Keras. The model was trained in a Jupyter notebook on a data scientist's local machine. The notebook contains a cell that performs data validation and a cell that performs model analysis. You need to orchestrate the steps contained in the notebook and automate the execution of these steps for weekly retraining. You expect much more training data in the future. You want your solution to take advantage of managed services while minimizing cost. What should you do?

Show Suggested Answer Hide Answer
Suggested Answer: D

by Beckie at Jan 09, 2023, 11:52 PM

Limited Time Offer

25%

Off

Get Premium Professional Machine Learning Engineer Questions as Interactive Web-Based Practice Test or PDF

Contribute your Thoughts:

0/2000 characters
Submit Cancel
Denna
9 months ago
D is interesting, but Airflow feels like overkill for this.
upvoted 0 times
...
An
9 months ago
C could work, but managing Spark jobs seems like a hassle.
upvoted 0 times
...
Carry
9 months ago
Wait, can we really automate everything with TFX? Sounds too good to be true!
upvoted 0 times
...
Raul
9 months ago
I think A is too costly for what we need.
upvoted 0 times
...
Carman
9 months ago
B is definitely the way to go for TFX and Vertex AI!
upvoted 0 times
...
Aliza
9 months ago
I vaguely recall something about scheduling with Cloud Scheduler, but I'm not convinced that A is the best approach for productionizing a model. It feels a bit too manual.
upvoted 0 times
...
Vincent
9 months ago
I practiced a similar question where we had to automate model retraining, and I think using Cloud Composer with Airflow could be a solid option, so D is appealing.
upvoted 0 times
...
Jesusita
9 months ago
I'm not entirely sure, but I feel like using Apache Spark could be overkill for this task. C seems like it might complicate things unnecessarily.
upvoted 0 times
...
Marcelle
10 months ago
I remember we discussed the importance of using managed services for scalability and cost-effectiveness, so I think B might be the right choice with TFX and Vertex AI.
upvoted 0 times
...
Joesph
10 months ago
Hmm, I'm a bit unsure about this one. I'll need to think through the key functions of Automated Change Management to determine the correct answers.
upvoted 0 times
...
Kenneth
10 months ago
Ugh, I'm a little unsure about this. Sprint duration is something we covered, but I'm having trouble remembering all the details. I'll have to think it through carefully.
upvoted 0 times
...
Caitlin
10 months ago
No problem, I've got this. The question is clear, and I know the module details, so I'm confident I can get this right.
upvoted 0 times
...
Sherita
10 months ago
Definitely seems like a 'True' answer - taking on work you're not qualified for is an ethics violation.
upvoted 0 times
...
Dorothy
1 year ago
I hope the exam doesn't ask us to 'productionize' a model using a typewriter and a fax machine. That would be a real challenge!
upvoted 0 times
...
Tracey
1 year ago
If this were a recipe, I'd say 'add a dash of TFX and a sprinkle of Vertex AI Pipelines for the perfect ML automation solution'.
upvoted 0 times
...
Sanda
1 year ago
Option A sounds like the easiest solution, but I'm worried about the cost of running a Notebooks instance on a large machine type. Managed services could be more cost-effective in the long run.
upvoted 0 times
Cathrine
1 year ago
D) Extract the steps contained in the Jupyter notebook as Python scripts, wrap each script in an Apache Airflow BashOperator, and run the resulting directed acyclic graph (DAG) in Cloud Composer.
upvoted 0 times
...
Maia
1 year ago
A) That's true, cost is definitely a concern. But using managed services like Vertex AI Pipelines could be more cost-effective in the long run.
upvoted 0 times
...
Willie
1 year ago
B) Write the code as a TensorFlow Extended (TFX) pipeline orchestrated with Vertex AI Pipelines. Use standard TFX components for data validation and model analysis, and use Vertex AI Pipelines for model retraining.
upvoted 0 times
...
Leigha
1 year ago
A) Move the Jupyter notebook to a Notebooks instance on the largest N2 machine type, and schedule the execution of the steps in the Notebooks instance using Cloud Scheduler.
upvoted 0 times
...
...
Kati
1 year ago
I'm not sure about option C. Rewriting the code as a Spark job might be overkill for this use case, and it might add unnecessary complexity.
upvoted 0 times
Alaine
1 year ago
D) Extract the steps contained in the Jupyter notebook as Python scripts, wrap each script in an Apache Airflow BashOperator, and run the resulting directed acyclic graph (DAG) in Cloud Composer.
upvoted 0 times
...
Alison
1 year ago
B) Write the code as a TensorFlow Extended (TFX) pipeline orchestrated with Vertex AI Pipelines. Use standard TFX components for data validation and model analysis, and use Vertex AI Pipelines for model retraining.
upvoted 0 times
...
Nickolas
1 year ago
A) Move the Jupyter notebook to a Notebooks instance on the largest N2 machine type, and schedule the execution of the steps in the Notebooks instance using Cloud Scheduler.
upvoted 0 times
...
...
Goldie
1 year ago
Option D looks interesting too. Airflow could provide a flexible way to orchestrate the different steps, and it's a popular tool in the industry.
upvoted 0 times
Nieves
1 year ago
Option D looks interesting too. Airflow could provide a flexible way to orchestrate the different steps, and it's a popular tool in the industry.
upvoted 0 times
...
Afton
1 year ago
D) Extract the steps contained in the Jupyter notebook as Python scripts, wrap each script in an Apache Airflow BashOperator, and run the resulting directed acyclic graph (DAG) in Cloud Composer.
upvoted 0 times
...
Donte
1 year ago
B) Write the code as a TensorFlow Extended (TFX) pipeline orchestrated with Vertex AI Pipelines. Use standard TFX components for data validation and model analysis, and use Vertex AI Pipelines for model retraining.
upvoted 0 times
...
...
Nichelle
1 year ago
I'm not sure about option B. I think option D could also work well by using Apache Airflow to orchestrate the steps in the Python scripts.
upvoted 0 times
...
Antonette
1 year ago
I agree with Jade. Option B seems like the most scalable and cost-effective solution for productionizing the ML model.
upvoted 0 times
...
Jade
1 year ago
I think option B is the best choice because using TFX pipeline with Vertex AI Pipelines will help automate the steps and handle the increasing amount of training data efficiently.
upvoted 0 times
...
Wilford
1 year ago
I'm leaning towards option B. Using TFX and Vertex AI Pipelines seems like a good way to take advantage of managed services and scale the solution as the data grows.
upvoted 0 times
Glen
1 year ago
I agree, using managed services like TFX and Vertex AI Pipelines will definitely help with scalability and cost-effectiveness in the long run.
upvoted 0 times
...
Floyd
1 year ago
Option B sounds like a solid choice. TFX and Vertex AI Pipelines can handle the increased data and automate the process efficiently.
upvoted 0 times
...
...

Save Cancel