Why scaling requires different strategies in MLOps - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to print the main reason why scaling needs different strategies.

MLOps

print('Scaling requires different strategies because of [1].')

Drag options to blanks, or click blank then click option'

Asingle user

Bsame hardware

Cfixed costs

Dvarying workloads

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing 'same hardware' which is not the main reason.

Thinking costs are fixed.

Scaling needs different strategies because workloads can vary a lot, so one size does not fit all.

2fill in blank

medium

Complete the sentence to explain a key challenge in scaling machine learning models.

MLOps

One challenge in scaling ML models is handling [1] data efficiently.

Drag options to blanks, or click blank then click option'

Alarge

Bstatic

Csmall

Dclean

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing 'small' which is opposite of the challenge.

Selecting 'clean' which is about data quality, not size.

Handling large data efficiently is a key challenge when scaling ML models.

3fill in blank

hard

Fix the error in the statement about scaling strategies.

MLOps

Scaling strategies must consider [1] and resource limits.

Drag options to blanks, or click blank then click option'

Auser interface

Bhardware uniformity

Cnetwork latency

Dsingle-threading

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing 'hardware uniformity' which is less relevant.

Selecting 'user interface' which is unrelated.

Network latency affects how scaling strategies perform, so it must be considered.

4fill in blank

hard

Fill both blanks to complete the scaling strategy code snippet.

MLOps

if workload [1] threshold:
    use [2] scaling

Drag options to blanks, or click blank then click option'

A>

Bhorizontal

Cvertical

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' which reverses the logic.

Choosing vertical scaling when workload is high.

If workload is greater than a threshold, horizontal scaling is used to add more machines.

5fill in blank

hard

Fill all three blanks to complete the dictionary comprehension for scaling decisions.

MLOps

scaling_decisions = {model: '[1]' if load [2] limit else '[3]' for model, load in loads.items()}

Drag options to blanks, or click blank then click option'

Ascale horizontally

B>

Cscale vertically

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' which reverses the condition.

Mixing up horizontal and vertical scaling.

If load is greater than limit, scale horizontally; otherwise, scale vertically.