Model registry in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to register a model named 'my_model' in MLflow. Assume 'model' is a scikit-learn model and registered_model_name='my_model' is used.

ML Python

import mlflow
mlflow.[1].log_model(model, 'my_model', registered_model_name='my_model')

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

Atracking

Bmodels

Cpytorch

Dsklearn

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'tracking' which handles experiments and tracking, not model logging.

Using 'models' or flavor-specific modules incorrectly.

Forgetting to import mlflow before using it.

The mlflow.sklearn module is used to log scikit-learn models to MLflow. Specifying registered_model_name='my_model' registers it in the model registry.

2fill in blank

medium

Complete the code to load a registered model named 'my_model' from MLflow.

ML Python

import mlflow
model = mlflow.pyfunc.load_model(model_uri='models:/[1]/1')

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

Amy_model

Bbest_model

Cmodel_v1

Dlatest_model

Attempts:

3 left

💡 Hint

Common Mistakes

Using a different model name than the one registered.

Forgetting the 'models:/' prefix in the URI.

The model URI must match the registered model name exactly, which is 'my_model' here.

3fill in blank

hard

Fix the error in the code to transition a model version to 'Production' stage.

ML Python

from mlflow.tracking import MlflowClient
client = MlflowClient()
client.[1](name='my_model', version=1, stage='Production')

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

Achange_stage

Btransition_model_version_stage

Cupdate_stage

Dset_model_stage

Attempts:

3 left

💡 Hint

Common Mistakes

Using incorrect method names that do not exist in MlflowClient.

Forgetting to import MlflowClient.

The correct method to change a model version's stage is transition_model_version_stage.

4fill in blank

hard

Fill both blanks to create a client and list all registered models.

ML Python

from mlflow.tracking import [1]
client = [2]() 
models = client.list_registered_models()

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

AMlflowClient

BMlflowRegistry

CModelClient

DModelRegistry

Attempts:

3 left

💡 Hint

Common Mistakes

Using non-existent classes like 'ModelClient' or 'MlflowRegistry'.

Importing the class but not instantiating it.

The MlflowClient class is used to interact with the model registry, including listing models.

5fill in blank

hard

Fill all three blanks to register a model, transition its stage, and load it.

ML Python

import mlflow
client = mlflow.tracking.[1]()
client.transition_model_version_stage(name='my_model', version=1, stage=[2])
model = mlflow.pyfunc.load_model(model_uri='models:/my_model/[3]')

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

AMlflowClient

B'Production'

DMlflowRegistry

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong client class names.

Not quoting the stage name correctly.

Using version numbers instead of stage names in the URI.

Use MlflowClient to transition the model stage to 'Production' and load the model from that stage.