Experiment tracking (MLflow) in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to start an MLflow experiment run.

ML Python

with mlflow.[1]():
    mlflow.log_param("param1", 5)

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

Astart_run

Brun_experiment

Cexperiment_start

Dbegin_run

Attempts:

3 left

💡 Hint

Common Mistakes

Using a non-existent function like mlflow.run_experiment()

Trying to call mlflow.experiment_start() which does not exist

The mlflow.start_run() function starts a new experiment run context.

2fill in blank

medium

Complete the code to log a metric value in MLflow.

ML Python

mlflow.[1]("accuracy", 0.95)

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

Alog_param

Blog_value

Clog_result

Dlog_metric

Attempts:

3 left

💡 Hint

Common Mistakes

Using mlflow.log_param() which is for parameters, not metrics

Trying to use mlflow.log_result() which does not exist

The mlflow.log_metric() function records a metric like accuracy during a run.

3fill in blank

hard

Fix the error in the code to set the experiment name in MLflow.

ML Python

mlflow.[1](experiment_name="MyExperiment")

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

Acreate_experiment

Bstart_experiment

Cset_experiment

Dexperiment_set

Attempts:

3 left

💡 Hint

Common Mistakes

Using mlflow.start_experiment() which does not exist

Trying mlflow.create_experiment() which requires an experiment name but is not for setting current experiment

The correct function to set or create an experiment by name is mlflow.set_experiment().

4fill in blank

hard

Fill both blanks to log a parameter and a metric inside an MLflow run.

ML Python

with mlflow.start_run():
    mlflow.[1]("learning_rate", 0.01)
    mlflow.[2]("rmse", 1.23)

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

Alog_param

Blog_metric

Clog_result

Dlog_value

Attempts:

3 left

💡 Hint

Common Mistakes

Using log_metric for parameters or log_param for metrics

Trying to use non-existent functions like log_result

Inside a run, use log_param to record parameters and log_metric to record metrics.

5fill in blank

hard

Fill all four blanks to create an MLflow run, log a parameter, a metric, and end the run.

ML Python

run = mlflow.[1]()
mlflow.[2]("batch_size", 32)
mlflow.[3]("accuracy", 0.89)
run.[4]()

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

Astart_run

Blog_param

Clog_metric

Dend_run

Attempts:

3 left

💡 Hint

Common Mistakes

Reusing start_run() to end the run instead of end_run()

Mixing up log_param and log_metric

Use mlflow.start_run() to begin a run (returns a run object), run.end_run() to end it, log_param() to log parameters, and log_metric() to log metrics.