将 AutoML 升级到 SDK v2

在 SDK v2 中，“试验”和“运行”合并到作业中。

在 SDK v1 中，主要使用 AutoMLConfig 类来配置和运行 AutoML。 在 SDK v2 中，此类已转换为 AutoML 作业。 尽管配置选项存在一些差异，但总体来说，命名和功能已在 V2 中保留。

本文比较 SDK v1 和 SDK v2 中的方案。

提交 AutoML 运行

• SDK v1：下面是一个示例 AutoML 分类任务。 有关整个代码，请查看示例存储库。

  # Imports
  
  import azureml.core
  from azureml.core.experiment import Experiment
  from azureml.core.workspace import Workspace
  from azureml.core.dataset import Dataset
  from azureml.train.automl import AutoMLConfig
  from azureml.train.automl.run import AutoMLRun   
  
  # Load tabular dataset
  data = "<url_to_data>"
  dataset = Dataset.Tabular.from_delimited_files(data)
  training_data, validation_data = dataset.random_split(percentage=0.8, seed=223)
  label_column_name = "Class"
  
  # Configure Auto ML settings
  automl_settings = {
      "n_cross_validations": 3,
      "primary_metric": "average_precision_score_weighted",
      "enable_early_stopping": True,
      "max_concurrent_iterations": 2,  
      "experiment_timeout_hours": 0.25,  
      "verbosity": logging.INFO,
  }
  
  # Put together an AutoML job constructor
  automl_config = AutoMLConfig(
      task="classification",
      debug_log="automl_errors.log",
      compute_target=compute_target,
      training_data=training_data,
      label_column_name=label_column_name,
      **automl_settings,
  )
  
  # Submit run
  remote_run = experiment.submit(automl_config, show_output=False)
  azureml_url = remote_run.get_portal_url()
  print(azureml_url)
  

• SDK v2：下面是一个示例 AutoML 分类任务。 有关整个代码，请查看示例存储库。

  # Imports
  from azure.ai.ml import automl, Input, MLClient
  
  from azure.ai.ml.constants import AssetTypes
  from azure.ai.ml.automl import (
      classification,
      ClassificationPrimaryMetrics,
      ClassificationModels,
  )
  
  
  # Create MLTables for training dataset
  # Note that AutoML Job can also take in tabular data
  my_training_data_input = Input(
      type=AssetTypes.MLTABLE, path="./data/training-mltable-folder"
  )
  
  # Create the AutoML classification job with the related factory-function.
  classification_job = automl.classification(
      compute="<compute_name>",
      experiment_name="<exp_name?",
      training_data=my_training_data_input,
      target_column_name="<name_of_target_column>",
      primary_metric="accuracy",
      n_cross_validations=5,
      enable_model_explainability=True,
      tags={"my_custom_tag": "My custom value"},
  )
  
  # Limits are all optional
  classification_job.set_limits(
      timeout_minutes=600,
      trial_timeout_minutes=20,
      max_trials=5,
      max_concurrent_trials = 4,
      max_cores_per_trial= 1,
      enable_early_termination=True,
  )
  
  # Training properties are optional
  classification_job.set_training(
      blocked_training_algorithms=["LogisticRegression"],
      enable_onnx_compatible_models=True,
  )
  
  # Submit the AutoML job
  returned_job = ml_client.jobs.create_or_update(classification_job)  
  returned_job
  

SDK v1 和 SDK v2 中关键功能的映射

后续步骤

有关详细信息，请参阅：

• 如何使用 Python SDKv2 训练 AutoML 模型