Online vs offline feature stores in MLOps - Interactive Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to specify the type of feature store used for real-time predictions.

MLOps

feature_store_type = "[1]"  # Used for serving features in real-time

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

Aoffline

Barchive

Conline

Dbatch

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing offline because it stores features, but it is for batch use.

The online feature store is used to serve features in real-time for predictions.

2fill in blank

medium

Complete the code to specify the feature store used for training machine learning models.

MLOps

training_feature_store = "[1]"  # Used for batch processing and model training

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

Astreaming

Bonline

Creal-time

Doffline

Attempts:

3 left

💡 Hint

Common Mistakes

Confusing online with offline because both store features.

The offline feature store is used for batch processing and training models with historical data.

3fill in blank

hard

Fix the error in the code by choosing the correct feature store type for low latency access.

MLOps

def get_feature_store(store_type):
    if store_type == "[1]":
        return "Low latency access for predictions"
    else:
        return "Batch access for training"

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

Aoffline

Bonline

Cbatch

Darchive

Attempts:

3 left

💡 Hint

Common Mistakes

Using offline instead of online for low latency.

The online feature store provides low latency access needed for real-time predictions.

4fill in blank

hard

Fill both blanks to create a dictionary mapping feature store types to their main use cases.

MLOps

feature_store_usage = {
    "[1]": "Used for batch training",
    "[2]": "Used for real-time serving"
}

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

Aoffline

Bonline

Carchive

Dstreaming

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing up offline and online in the dictionary.

The offline store is for batch training, and the online store is for real-time serving.

5fill in blank

hard

Fill all three blanks to complete the code that checks feature store latency and usage.

MLOps

def check_store(store):
    latency = store.get_latency()
    if latency [1] 10:
        usage = "[2]"
    else:
        usage = "[3]"
    return usage

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

A<=

Bonline

Coffline

D>

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong comparison operator or swapping online/offline usage.

If latency is less or equal to 10 ms, the store is online (real-time). Otherwise, it is offline (batch).