Merge pull request Azure#1211 from Azure/release_update/Release-72

update samples from Release-72 as a part of  SDK release

Author: harneetvirk

README.md

@@ -1,7 +1,5 @@
 # Azure Machine Learning service example notebooks
 
-> A community-driven repository of training and scoring examples can be found at https://github.com/Azure/azureml-examples
-
 This repository contains example notebooks demonstrating the [Azure Machine Learning](https://azure.microsoft.com/en-us/services/machine-learning-service/) Python SDK which allows you to build, train, deploy and manage machine learning solutions using Azure.  The AML SDK allows you the choice of using local or cloud compute resources, while managing and maintaining the complete data science workflow from the cloud.
 
 ![Azure ML Workflow](https://raw.githubusercontent.com/MicrosoftDocs/azure-docs/master/articles/machine-learning/media/concept-azure-machine-learning-architecture/workflow.png)

how-to-use-azureml/automated-machine-learning/forecasting-beer-remote/infer.py

@@ -10,6 +10,8 @@
 from azureml.automl.runtime.shared.score import scoring, constants
 from azureml.core import Run
 
+import torch
+
 
 def align_outputs(y_predicted, X_trans, X_test, y_test,
                   predicted_column_name='predicted',
@@ -221,6 +223,10 @@ def MAPE(actual, pred):
     return np.mean(APE(actual_safe, pred_safe))
 
 
+def map_location_cuda(storage, loc):
+    return storage.cuda()
+
+
 parser = argparse.ArgumentParser()
 parser.add_argument(
     '--max_horizon', type=int, dest='max_horizon',
@@ -274,8 +280,13 @@ def MAPE(actual, pred):
 y_lookback_df = lookback_dataset.with_timestamp_columns(
     None).keep_columns(columns=[target_column_name])
 
-fitted_model = joblib.load(model_path)
-
+# Load the trained model with torch.
+if torch.cuda.is_available():
+    map_location = map_location_cuda
+else:
+    map_location = 'cpu'
+with open(model_path, 'rb') as fh:
+    fitted_model = torch.load(fh, map_location=map_location)
 
 if hasattr(fitted_model, 'get_lookback'):
     lookback = fitted_model.get_lookback()