Merge pull request #115912 from nibaccam/ensemble-param

AutoML | Ensemble param

Author: PRMerger19

articles/machine-learning/how-to-configure-auto-train.md

@@ -4,12 +4,12 @@ titleSuffix: Azure Machine Learning
 description: Automated machine learning picks an algorithm for you and generates a model ready for deployment. Learn the options that you can use to configure automated machine learning experiments.
 author: cartacioS
 ms.author: sacartac
-ms.reviewer: sgilley
+ms.reviewer: nibaccam
 services: machine-learning
 ms.service: machine-learning
 ms.subservice: core
 ms.topic: conceptual
-ms.date: 03/09/2020
+ms.date: 05/20/2020
 ms.custom: seodec18
 ---
 
@@ -249,16 +249,23 @@ automl_config = AutoMLConfig(task = 'forecasting',
 
 Ensemble models are enabled by default, and appear as the final run iterations in an automated machine learning run. Currently supported ensemble methods are voting and stacking. Voting is implemented as soft-voting using weighted averages, and the stacking implementation is using a two layer implementation, where the first layer has the same models as the voting ensemble, and the second layer model is used to find the optimal combination of the models from the first layer. If you are using ONNX models, **or** have model-explainability enabled, stacking will be disabled and only voting will be utilized.
 
-There are multiple default arguments that can be provided as `kwargs` in an `AutoMLConfig` object to alter the default stack ensemble behavior.
+There are multiple default arguments that can be provided as `kwargs` in an `AutoMLConfig` object to alter the default ensemble behavior.
+
+* `ensemble_download_models_timeout_sec`: During VotingEnsemble and StackEnsemble model generation, multiple fitted models from the previous child runs are downloaded. If you encounter this error: `AutoMLEnsembleException: Could not find any models for running ensembling`, then you may need to provide more time for the models to be downloaded. The default value is 300 seconds for downloading these models in parallel and there is no maximum timeout limit. Configure this parameter with a higher value than 300 secs, if more time is needed. **Note**: If the timeout is reached and there are models downloaded, then the ensembling proceeds with as many models it has downloaded (it's not required that all the models need to be downloaded to finish within that timeout).
+
+The following parameters only apply to StackEnsemble models: 
 
 * `stack_meta_learner_type`: the meta-learner is a model trained on the output of the individual heterogeneous models. Default meta-learners are `LogisticRegression` for classification tasks (or `LogisticRegressionCV` if cross-validation is enabled) and `ElasticNet` for regression/forecasting tasks (or `ElasticNetCV` if cross-validation is enabled). This parameter can be one of the following strings: `LogisticRegression`, `LogisticRegressionCV`, `LightGBMClassifier`, `ElasticNet`, `ElasticNetCV`, `LightGBMRegressor`, or `LinearRegression`.
-* `stack_meta_learner_train_percentage`: specifies the proportion of the training set (when choosing train and validation type of training) to be reserved for training the meta-learner. Default value is `0.2`.
+
+* `stack_meta_learner_train_percentage`: specifies the proportion of the training set (when choosing train and validation type of training) to be reserved for training the meta-learner. Default value is `0.2`. 
+
 * `stack_meta_learner_kwargs`: optional parameters to pass to the initializer of the meta-learner. These parameters and parameter types mirror the parameters and parameter types from the corresponding model constructor, and are forwarded to the model constructor.
 
 The following code shows an example of specifying custom ensemble behavior in an `AutoMLConfig` object.
 
 ```python
 ensemble_settings = {
+    "ensemble_download_models_timeout_sec": 600
     "stack_meta_learner_type": "LogisticRegressionCV",
     "stack_meta_learner_train_percentage": 0.3,
     "stack_meta_learner_kwargs": {