Cost allocation and optimization in MLOps - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to set a budget alert threshold at 80%.

MLOps

budget_alert = BudgetAlert(threshold= [1])

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

A0.8

B80

C8

D1.2

Attempts:

3 left

💡 Hint

Common Mistakes

Using 80 instead of 0.8

Using values greater than 1

The threshold should be a decimal representing 80%, so 0.8 is correct.

2fill in blank

medium

Complete the code to filter cost data for the current month.

MLOps

filtered_costs = costs.filter(date__month= [1])

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

A12

Bcurrent_month

C30

Ddatetime.now().month

Attempts:

3 left

💡 Hint

Common Mistakes

Using a string instead of an integer

Hardcoding the month number

Using datetime.now().month dynamically gets the current month number.

3fill in blank

hard

Fix the error in the code to calculate cost per model deployment.

MLOps

cost_per_deployment = total_cost [1] number_of_deployments

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

A/

B*

C+

D-

Attempts:

3 left

💡 Hint

Common Mistakes

Using multiplication instead of division

Using addition or subtraction

To find cost per deployment, divide total cost by number of deployments.

4fill in blank

hard

Fill both blanks to create a dictionary of model costs filtered by cost greater than 100.

MLOps

model_costs = {model: cost [1] model, cost in cost_data.items() if cost [2] 100}

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

Afor

B==

C>

Din

Attempts:

3 left

💡 Hint

Common Mistakes

Using '==' instead of '>' for filtering

Using 'in' instead of 'for' in comprehension

The dictionary comprehension needs 'for' to iterate and '>' to filter costs above 100.

5fill in blank

hard

Fill all three blanks to create a filtered dictionary of model costs where cost is less than 500 and keys are uppercase.

MLOps

filtered_costs = [1]: cost for [2], cost in costs.items() if cost [3] 500

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

Amodel.upper()

Bmodel

C<

Dcost

Attempts:

3 left

💡 Hint

Common Mistakes

Using cost as key instead of model

Using '>' instead of '<' for filtering

Keys are converted to uppercase with model.upper(), iterated with model, and filtered with <.

Practice

1. What is the main purpose of cost allocation in MLOps?

easy

A. To improve model accuracy

B. To increase the speed of model training

C. To track who uses resources and how much they cost

D. To automate data labeling

Solution

1. Step 1: Understand cost allocation concept

   Cost allocation means assigning costs to users or projects to see usage and expenses clearly.

2. Step 2: Identify the main goal in MLOps

   In MLOps, cost allocation helps track resource usage and spending by teams or projects.

3. Final Answer:

   To track who uses resources and how much they cost -> Option C

4. Quick Check:

   Cost allocation = track usage and cost [OK]

Hint: Cost allocation = who uses what and cost [OK]

Common Mistakes:

• Confusing cost allocation with model accuracy
• Thinking cost allocation speeds up training
• Mixing cost allocation with automation tasks

2. Which of the following is the correct syntax to tag a resource for cost allocation in a YAML MLOps config?

easy

A. tags: [owner=team-alpha, project=fraud-detection]

B. tags = {owner: team-alpha, project: fraud-detection}

C. tags: owner: team-alpha; project: fraud-detection

D. tags:\n owner: team-alpha\n project: fraud-detection

Solution

1. Step 1: Recognize YAML syntax for key-value pairs

   YAML uses colon and indentation for mapping keys to values, like 'tags:\n owner: value'.

2. Step 2: Compare options to YAML format

   tags:\n owner: team-alpha\n project: fraud-detection uses correct YAML indentation and colon syntax for tags; others use invalid syntax.

3. Final Answer:

   tags:\n owner: team-alpha\n project: fraud-detection -> Option D

4. Quick Check:

   YAML tags use colon and indentation [OK]

Hint: YAML uses colon and indentation for tags [OK]

Common Mistakes:

• Using equal signs instead of colons in YAML
• Putting tags in brackets like a list
• Separating tags with semicolons

3. Given this Python snippet for cost optimization, what is the output?

costs = [100, 200, 300, 400]
optimized = [c * 0.8 for c in costs if c > 150]
print(optimized)

medium

A. [80.0, 160.0, 240.0, 320.0]

B. [160.0, 240.0, 320.0]

C. [200, 300, 400]

D. [80, 160, 240]

Solution

1. Step 1: Filter costs greater than 150

   From the list, values > 150 are 200, 300, 400.

2. Step 2: Apply 20% discount (multiply by 0.8)

   200*0.8=160.0, 300*0.8=240.0, 400*0.8=320.0.

3. Final Answer:

   [160.0, 240.0, 320.0] -> Option B

4. Quick Check:

   Filter >150 then multiply by 0.8 = [160.0, 240.0, 320.0] [OK]

Hint: Filter costs >150 then multiply by 0.8 [OK]

Common Mistakes:

• Applying discount to all costs instead of filtered
• Forgetting to filter costs >150
• Using integer instead of float multiplication

4. You have this snippet to tag resources but it causes an error:

tags:
  owner: team-alpha
  project fraud-detection

What is the error and how to fix it?

medium

A. Missing colon after 'project'; fix by adding ':' like 'project: fraud-detection'

B. Wrong indentation; fix by indenting 'project' more

C. Tags must be in quotes; fix by adding quotes around values

D. Use equal sign instead of colon; fix by 'project = fraud-detection'

Solution

1. Step 1: Identify YAML syntax error

   YAML requires a colon ':' after keys; 'project fraud-detection' misses the colon.

2. Step 2: Correct the syntax

   Add colon after 'project' to become 'project: fraud-detection' to fix error.

3. Final Answer:

   Missing colon after 'project'; fix by adding ':' like 'project: fraud-detection' -> Option A

4. Quick Check:

   YAML keys need colon ':' [OK]

Hint: YAML keys must end with colon ':' [OK]

Common Mistakes:

• Ignoring missing colon errors
• Changing indentation instead of fixing colon
• Using equal signs in YAML

5. You want to optimize costs by automatically stopping idle compute instances after 30 minutes. Which approach combines cost allocation and optimization best?

hard

A. Tag instances by owner and project, then use a script to stop idle instances after 30 minutes

B. Only tag instances by owner without automation

C. Manually check instances daily and stop idle ones

D. Increase instance size to reduce runtime

Solution

1. Step 1: Use cost allocation tags

   Tagging by owner and project helps track who uses which resources and their costs.

2. Step 2: Automate cost optimization

   Using a script to stop idle instances after 30 minutes saves money by reducing waste.

3. Step 3: Combine both for best results

   Tagging plus automation ensures clear cost tracking and efficient spending control.

4. Final Answer:

   Tag instances by owner and project, then use a script to stop idle instances after 30 minutes -> Option A

5. Quick Check:

   Tag + automate stopping idle = best cost control [OK]

Hint: Combine tagging with automation to save costs [OK]

Common Mistakes:

• Skipping automation and relying on manual checks
• Tagging without any optimization steps
• Increasing instance size without cost control