Why experiment tracking prevents wasted work in MLOps - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to start tracking an experiment run using MLflow.

MLOps

with mlflow.start_run() as [1]:
    mlflow.log_param("learning_rate", 0.01)

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

Atracker

Bexperiment

Csession

Drun

Attempts:

3 left

💡 Hint

Common Mistakes

Using unrelated variable names like 'experiment' or 'session' which do not represent the run context.

The variable run is commonly used to represent the active experiment run context in MLflow.

2fill in blank

medium

Complete the code to log a metric called 'accuracy' with value 0.95 in MLflow.

MLOps

mlflow.log_metric("accuracy", [1])

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

A0.95

B0.85

C1.0

D0.75

Attempts:

3 left

💡 Hint

Common Mistakes

Logging incorrect metric values like 0.85 or 1.0 which do not match the intended accuracy.

The metric value 0.95 matches the accuracy we want to log.

3fill in blank

hard

Fix the error in the code to set the experiment name before starting a run.

MLOps

mlflow.[1]("MyExperiment")
with mlflow.start_run():
    mlflow.log_param("batch_size", 32)

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

Aset_experiment_name

Bcreate_experiment

Cset_experiment

Dstart_experiment

Attempts:

3 left

💡 Hint

Common Mistakes

Using non-existent functions like 'set_experiment_name' or 'start_experiment'.

The correct MLflow function to set the experiment name is set_experiment.

4fill in blank

hard

Fill both blanks to create a dictionary of parameters and log them in MLflow.

MLOps

params = {"epochs": [1], "dropout": [2]
mlflow.log_params(params)

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

A10

B0.3

C5

D0.5

Attempts:

3 left

💡 Hint

Common Mistakes

Swapping values or using incorrect types like strings instead of numbers.

The parameters dictionary uses 10 epochs and 0.3 dropout rate as values.

5fill in blank

hard

Fill all three blanks to filter experiment runs with accuracy greater than 0.9.

MLOps

runs = mlflow.search_runs(filter_string="metrics.accuracy [1] [2]")
high_acc_runs = [run for run in runs.itertuples() if run.metrics_accuracy [3] 0.9]

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

A>=

B>

C0.9

D0.8

Attempts:

3 left

💡 Hint

Common Mistakes

Using '>=' or wrong threshold values like 0.8 which change the filter meaning.

The filter string uses '>' and the threshold 0.9 to find runs with accuracy greater than 0.9.