Audit trails for model decisions in MLOps - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to log model predictions to a file.

MLOps

with open('audit_log.txt', '[1]') as f:
    f.write(f"Prediction: {prediction}\n")

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

Ar

Ba

Cw+

Dx

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'w' which overwrites the file, losing previous logs.

Use 'a' to append logs without overwriting existing data.

2fill in blank

medium

Complete the code to include a timestamp in the audit log entry.

MLOps

import datetime

log_entry = f"{datetime.datetime.now()}: Prediction = {prediction}"
with open('audit_log.txt', 'a') as f:
    f.write([1] + '\n')

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

Alog_entry()

Bstr(log_entry)

Crepr(log_entry)

Dlog_entry

Attempts:

3 left

💡 Hint

Common Mistakes

Calling log_entry() which causes an error because it's not a function.

log_entry is already a string, so write it directly.

3fill in blank

hard

Fix the error in the code that logs model input and output as JSON.

MLOps

import json

log_data = {'input': model_input, 'output': model_output}
with open('audit_log.json', 'a') as f:
    f.write(json.dumps([1]) + '\n')

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

Astr(log_data)

Blog_data()

Clog_data

Djson.dumps(log_data)

Attempts:

3 left

💡 Hint

Common Mistakes

Passing log_data() which causes a TypeError.

json.dumps expects a dictionary, so pass log_data directly.

4fill in blank

hard

Fill both blanks to create a dictionary comprehension that logs only features with values above 0.5.

MLOps

filtered_log = {key: value for key, value in model_features.items() if value [1] 0.5 and key [2] 'confidence'}

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

A>

B<

C!=

D==

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' instead of '>' for value comparison.

Using '==' instead of '!=' for key filtering.

We want values greater than 0.5 and keys not equal to 'confidence'.

5fill in blank

hard

Fill all three blanks to create a filtered audit log dictionary with uppercase keys, values above 0.7, and keys not equal to 'bias'.

MLOps

filtered_audit = [1]: [2] for [3], val in audit_data.items() if val > 0.7 and [3] != 'bias'

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

Akey.upper()

Bval

Ckey

Dvalue

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'value' instead of 'val' for values.

Using 'value' or 'val' as loop variable instead of 'key'.

Use key.upper() for keys, val for values, and key as the loop variable.