MLflow setup and basics in MLOps - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to start the MLflow tracking server locally.

MLOps

mlflow [1]

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

Astart

Brun

Cserver

Dui

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'run' instead of 'ui' starts an MLflow project run, not the server.

Using 'server' requires --backend-store-uri and other flags; 'start' is not a valid MLflow command.

The command mlflow ui starts the MLflow tracking server locally with a web UI.

2fill in blank

medium

Complete the code to log a parameter named 'alpha' with value 0.5 in MLflow.

MLOps

mlflow.log_param('[1]', 0.5)

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

Alearning_rate

Blambda

Calpha

Dbeta

Attempts:

3 left

💡 Hint

Common Mistakes

Using other parameter names like 'beta' or 'lambda' instead of 'alpha'.

Forgetting to put the parameter name in quotes.

The parameter name to log is 'alpha' as specified.

3fill in blank

hard

Fix the error in the code to start an MLflow run context.

MLOps

with mlflow.[1]():
    print('Running experiment')

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

Astart_run

Brun_start

Cbegin_run

Drun

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'run' or 'begin_run' which are not valid MLflow methods.

Swapping the order of words in the method name.

The correct method to start a run context in MLflow is start_run().

4fill in blank

hard

Fill both blanks to log a metric named 'accuracy' with value 0.95 inside an MLflow run.

MLOps

with mlflow.[1]():
    mlflow.[2]('accuracy', 0.95)

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

Astart_run

Blog_metric

Clog_param

Drun

Attempts:

3 left

💡 Hint

Common Mistakes

Using log_param instead of log_metric for metrics.

Using run instead of start_run to start the run.

Use start_run() to begin the run and log_metric() to log the metric.

5fill in blank

hard

Fill all three blanks to create an MLflow experiment named 'MyExperiment' and set it as active.

MLOps

mlflow.[1]('MyExperiment')
experiment_id = mlflow.[2]('MyExperiment').experiment_id
mlflow.[3](experiment_id)

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

Acreate_experiment

Bget_experiment_by_name

Cset_experiment

Dstart_run

Attempts:

3 left

💡 Hint

Common Mistakes

Using start_run instead of create_experiment or set_experiment.

Confusing get_experiment_by_name with set_experiment.

First, create the experiment with create_experiment. Then get its info with get_experiment_by_name. Finally, set it active with set_experiment.