Pyspark ML:如何使用CrossValidator()获取subModels值

from pyspark.ml.tuning import CrossValidator, ParamGridBuilder
from pyspark.ml.evaluation import BinaryClassificationEvaluator

from pyspark.ml.classification import LogisticRegression

lr = LogisticRegression()

paramGrid = ParamGridBuilder() \
    .addGrid(lr.regParam, [0.01, 0.1, 1.0]) \
    .addGrid(lr.elasticNetParam, [0.0, 0.5, 1.0]) \
    .build()

evaluator = BinaryClassificationEvaluator()

crossval = CrossValidator(estimator=lr,
                          estimatorParamMaps=paramGrid,
                          evaluator=evaluator,
                          numFolds=3)

cvModel = crossval.fit(train_data)

bestModel = cvModel.bestModel
bestParams = bestModel.extractParamMap()

subModels = cvModel.subModels
for i, model in enumerate(subModels):
    params = model.extractParamMap()
    score = evaluator.evaluate(model.transform(test_data))
    print("Sub-model", i+1, "params:", params)
    print("Sub-model", i+1, "score:", score)