how and where to edit enable_early_stopping = False in AutoMLConfig for notebook/python SDK runs

Hello Paromita Biswas,

Welcome to the Microsoft Q&A and thank you for posting your questions here.

Problem

I understand that you have a challenge to enable_early_stopping = False, in AutoMLConfig for notebook/python SDK runs. Also, you would like to know how and where to solve this issue.

Solution

There are two things based on practice, if you are using old notebook and Python SDK and it requested you to create AutoMLConfig. You will create this by yourself and ensure you are able to enable_early_stopping = False This is how to do it:

# Import necessary Libraries
from azureml.core.workspace import Workspace
from azureml.train.automl import AutoMLConfig
from azureml.core.experiment import Experiment
# Load the workspace
ws = Workspace.from_config()
# Create AutoMLConfig
automl_config = AutoMLConfig(
    task='classification',
    primary_metric='accuracy',
    experiment_timeout_minutes=30,
    training_data=train_data,  # Replace with your dataset
    label_column_name='target',  # Replace with your label column
    n_cross_validations=5,
    enable_early_stopping=False,
    # Add other parameters as needed
)
# Run the experiment
experiment = Experiment(ws, 'your_experiment_name')
run = experiment.submit(automl_config, show_output=True)

Secondly, if you have updated version, AutoMLConfig has been replaced with a new configuration object in the Azure Machine Learning Python SDK, specifically AutoMLJob within the azure.ai.ml package. Especially, in a Jupyter notebook or any Python environment, you can disable early stopping with the new AutoMLJob class by following these steps:

1. Ensure you have the latest Azure ML SDK installed using bash: pip install azure-ai-ml

2.    # Import necessary Libraries
      from azure.ai.ml import automl
      from azure.ai.ml.entities import AutomlClassificationJob
      from azure.ai.ml import MLClient
      from azure.identity import DefaultAzureCredential
      # Load the workspace
      ml_client = MLClient(
          DefaultAzureCredential(),
          subscription_id="your-subscription-id",
          resource_group_name="your-resource-group",
          workspace_name="your-workspace-name"
      )
      # Create AutoML Job Configuration
      classification_job = automl.classification(
          compute="your-compute-cluster",
          experiment_name="your-experiment-name",
          training_data="path/to/your/training/data",
          label_column_name="your-label-column",
          primary_metric="accuracy",
          n_cross_validations=5,
          enable_early_stopping=False,  # Disable early stopping
          # Add other necessary parameters
      )
      # Submit the job
      returned_job = ml_client.jobs.create_or_update(classification_job)
      ml_client.jobs.stream(returned_job.name)
   

References

Kindly read more from the additional resources available by the right side of this page, and utilize the following links for more details.

Source: AutoMLConfig class. Accessed, 6/25/2024.

Source: Automated Machine Learning for Classification -Notebook examples. Accessed, 6/25/2024.

Source: Automated ML Jobs. Accessed, 6/25/2024.

Source: Azure ML SDK Documentation for azure-ai-ml package. Accessed, 6/25/2024.

Accept Answer

I hope this is helpful! Do not hesitate to let me know if you have any other questions.

** Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful ** so that others in the community facing similar issues can easily find the solution.

Best Regards,

Sina Salam