Azure · harneetvirk · Feb 11, 2020 · Feb 11, 2020
diff --git a/configuration.ipynb b/configuration.ipynb
@@ -103,7 +103,7 @@
       "source": [
         "import azureml.core\n",
         "\n",
-        "print(\"This notebook was created using version 1.1.0rc0 of the Azure ML SDK\")\n",
+        "print(\"This notebook was created using version 1.1.0rc1 of the Azure ML SDK\")\n",
         "print(\"You are currently using version\", azureml.core.VERSION, \"of the Azure ML SDK\")"
       ]
     },

diff --git a/how-to-use-azureml/automated-machine-learning/automl_env.yml b/how-to-use-azureml/automated-machine-learning/automl_env.yml
@@ -23,7 +23,6 @@ dependencies:
   - azureml-train-automl
   - azureml-train
   - azureml-widgets
-  - azureml-explain-model
   - azureml-pipeline
   - azureml-contrib-interpret
   - pytorch-transformers==1.0.0
@@ -34,4 +33,4 @@ dependencies:
 
 channels:
 - conda-forge
-- pytorch
+- pytorch
diff --git a/how-to-use-azureml/automated-machine-learning/automl_env_mac.yml b/how-to-use-azureml/automated-machine-learning/automl_env_mac.yml
@@ -24,7 +24,6 @@ dependencies:
   - azureml-train-automl
   - azureml-train
   - azureml-widgets
-  - azureml-explain-model
   - azureml-pipeline
   - azureml-contrib-interpret
   - pytorch-transformers==1.0.0
@@ -35,4 +34,4 @@ dependencies:
 
 channels:
 - conda-forge
-- pytorch
+- pytorch
diff --git a/...-learning/classification-credit-card-fraud/auto-ml-classification-credit-card-fraud.ipynb b/...-learning/classification-credit-card-fraud/auto-ml-classification-credit-card-fraud.ipynb
@@ -212,7 +212,7 @@
         "    \"primary_metric\": 'average_precision_score_weighted',\n",
         "    \"enable_early_stopping\": True,\n",
         "    \"max_concurrent_iterations\": 2, # This is a limit for testing purpose, please increase it as per cluster size\n",
-        "    \"experiment_timeout_hours\": 0.2, # This is a time limit for testing purposes, remove it for real use cases, this will drastically limit ablity to find the best model possible\n",
+        "    \"experiment_timeout_hours\": 0.25, # This is a time limit for testing purposes, remove it for real use cases, this will drastically limit ablity to find the best model possible\n",
         "    \"verbosity\": logging.INFO,\n",
         "}\n",
         "\n",

diff --git a/.../automated-machine-learning/classification-text-dnn/auto-ml-classification-text-dnn.ipynb b/.../automated-machine-learning/classification-text-dnn/auto-ml-classification-text-dnn.ipynb
@@ -121,9 +121,9 @@
       "metadata": {},
       "source": [
         "## Set up a compute cluster\n",
-        "This section uses a user-provided compute cluster (named \"cpu-cluster\" in this example). If a cluster with this name does not exist in the user's workspace, the below code will create a new cluster. You can choose the parameters of the cluster as mentioned in the comments.\n",
+        "This section uses a user-provided compute cluster (named \"dnntext-cluster\" in this example). If a cluster with this name does not exist in the user's workspace, the below code will create a new cluster. You can choose the parameters of the cluster as mentioned in the comments.\n",
         "\n",
-        "Whether you provide/select a CPU or GPU cluster, AutoML will choose the appropriate DNN for that setup - BiLSTM or BERT text featurizer will be included in the candidate featurizers on CPU and GPU respectively."
+        "Whether you provide/select a CPU or GPU cluster, AutoML will choose the appropriate DNN for that setup - BiLSTM or BERT text featurizer will be included in the candidate featurizers on CPU and GPU respectively.  If your goal is to obtain the most accurate model, we recommend you use GPU clusters since BERT featurizers usually outperform BiLSTM featurizers."
       ]
     },
     {
@@ -133,7 +133,7 @@
       "outputs": [],
       "source": [
         "# Choose a name for your cluster.\n",
-        "amlcompute_cluster_name = \"cpu-dnntext\"\n",
+        "amlcompute_cluster_name = \"dnntext-cluster\"\n",
         "\n",
         "found = False\n",
         "# Check if this compute target already exists in the workspace.\n",
@@ -145,11 +145,11 @@
         "\n",
         "if not found:\n",
         "    print('Creating a new compute target...')\n",
-        "    provisioning_config = AmlCompute.provisioning_configuration(vm_size = \"STANDARD_D2_V2\", # CPU for BiLSTM\n",
-        "                                                                # To use BERT, select a GPU such as \"STANDARD_NC6\" \n",
+        "    provisioning_config = AmlCompute.provisioning_configuration(vm_size = \"STANDARD_NC6\", # CPU for BiLSTM, such as \"STANDARD_D2_V2\" \n",
+        "                                                                # To use BERT (this is recommended for best performance), select a GPU such as \"STANDARD_NC6\" \n",
         "                                                                # or similar GPU option\n",
         "                                                                # available in your workspace\n",
-        "                                                                max_nodes = 6)\n",
+        "                                                                max_nodes = 1)\n",
         "\n",
         "    # Create the cluster\n",
         "    compute_target = ComputeTarget.create(ws, amlcompute_cluster_name, provisioning_config)\n",
@@ -218,7 +218,7 @@
       "cell_type": "markdown",
       "metadata": {},
       "source": [
-        "Featch data and upload to datastore for use in training"
+        "#### Fetch data and upload to datastore for use in training"
       ]
     },
     {
@@ -347,7 +347,26 @@
       "metadata": {},
       "outputs": [],
       "source": [
-        "#best_run, fitted_model = automl_run.get_output()"
+        "best_run, fitted_model = automl_run.get_output()"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "metadata": {},
+      "source": [
+        "You can now see what text transformations are used to convert text data to features for this dataset, including deep learning transformations based on BiLSTM or Transformer (BERT is one implementation of a Transformer) models."
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "metadata": {},
+      "outputs": [],
+      "source": [
+        "text_transformations_used = []\n",
+        "for column_group in fitted_model.named_steps['datatransformer'].get_featurization_summary():\n",
+        "    text_transformations_used.extend(column_group['Transformations'])\n",
+        "text_transformations_used"
       ]
     },
     {

diff --git a/...ml/automated-machine-learning/classification-text-dnn/auto-ml-classification-text-dnn.yml b/...ml/automated-machine-learning/classification-text-dnn/auto-ml-classification-text-dnn.yml
@@ -6,3 +6,8 @@ dependencies:
   - azureml-widgets
   - matplotlib
   - azurmel-train
+  - https://download.pytorch.org/whl/cpu/torch-1.1.0-cp35-cp35m-win_amd64.whl
+  - sentencepiece==0.1.82
+  - pytorch-transformers==1.0
+  - spacy==2.1.8
+  - https://aka.ms/automl-resources/packages/en_core_web_sm-2.1.0.tar.gz
diff --git a/...reml/automated-machine-learning/continuous-retraining/auto-ml-continuous-retraining.ipynb b/...reml/automated-machine-learning/continuous-retraining/auto-ml-continuous-retraining.ipynb
@@ -197,7 +197,7 @@
         "conda_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
         "\n",
         "cd = CondaDependencies.create(pip_packages=['azureml-sdk[automl]', 'applicationinsights', 'azureml-opendatasets'], \n",
-        "                              conda_packages=['numpy', 'py-xgboost'], \n",
+        "                              conda_packages=['numpy==1.16.2'], \n",
         "                              pin_sdk_version=False)\n",
         "#cd.add_pip_package('azureml-explain-model')\n",
         "conda_run_config.environment.python.conda_dependencies = cd\n",
@@ -277,7 +277,7 @@
       "metadata": {},
       "outputs": [],
       "source": [
-        "data_pipeline_run.wait_for_completion()"
+        "data_pipeline_run.wait_for_completion(show_output=False)"
       ]
     },
     {
@@ -347,7 +347,7 @@
         "\n",
         "automl_settings = {\n",
         "    \"iteration_timeout_minutes\": 10,\n",
-        "    \"experiment_timeout_hours\": 0.2,\n",
+        "    \"experiment_timeout_hours\": 0.25,\n",
         "    \"n_cross_validations\": 3,\n",
         "    \"primary_metric\": 'r2_score',\n",
         "    \"max_concurrent_iterations\": 3,\n",

diff --git a/...run-classification-credit-card-fraud/auto-ml-classification-credit-card-fraud-local.ipynb b/...run-classification-credit-card-fraud/auto-ml-classification-credit-card-fraud-local.ipynb
@@ -155,7 +155,7 @@
         "automl_settings = {\n",
         "    \"n_cross_validations\": 3,\n",
         "    \"primary_metric\": 'average_precision_score_weighted',\n",
-        "    \"experiment_timeout_hours\": 0.2, # This is a time limit for testing purposes, remove it for real use cases, this will drastically limit ability to find the best model possible\n",
+        "    \"experiment_timeout_hours\": 0.25, # This is a time limit for testing purposes, remove it for real use cases, this will drastically limit ability to find the best model possible\n",
         "    \"verbosity\": logging.INFO,\n",
         "    \"enable_stack_ensemble\": False\n",
         "}\n",

diff --git a/...featurization/auto-ml-regression-hardware-performance-explanation-and-featurization.ipynb b/...featurization/auto-ml-regression-hardware-performance-explanation-and-featurization.ipynb
@@ -262,7 +262,7 @@
       "source": [
         "automl_settings = {\n",
         "    \"enable_early_stopping\": True, \n",
-        "    \"experiment_timeout_hours\" : 0.2,\n",
+        "    \"experiment_timeout_hours\" : 0.25,\n",
         "    \"max_concurrent_iterations\": 4,\n",
         "    \"max_cores_per_iteration\": -1,\n",
         "    \"n_cross_validations\": 5,\n",

diff --git a/how-to-use-azureml/azureml-sdk-for-r/README.md b/how-to-use-azureml/azureml-sdk-for-r/README.md
@@ -0,0 +1,36 @@
+## Examples to get started with Azure Machine Learning SDK for R
+
+Learn how to use Azure Machine Learning SDK for R for experimentation and model management.
+
+As a pre-requisite, go through the [Installation](vignettes/installation.Rmd) and [Configuration](vignettes/configuration.Rmd) vignettes to first install the package and set up your Azure Machine Learning Workspace unless you are running these examples on an Azure Machine Learning compute instance. Azure Machine Learning compute instances have the Azure Machine Learning SDK pre-installed and your workspace details pre-configured.
+
+
+Samples
+* Deployment
+	* [deploy-to-aci](./samples/deployment/deploy-to-aci): Deploy a model as a web service to Azure Container Instances (ACI).
+	* [deploy-to-local](./samples/deployment/deploy-to-local): Deploy a model as a web service locally.
+* Training
+	* [train-on-amlcompute](./samples/training/train-on-amlcompute): Train a model on a remote AmlCompute cluster.
+	* [train-on-local](./samples/training/train-on-local): Train a model locally with Docker.
+
+Vignettes
+* [deploy-to-aks](./vignettes/deploy-to-aks): Production deploy a model as a web service to Azure Kubernetes Service (AKS).
+* [hyperparameter-tune-with-keras](./vignettes/hyperparameter-tune-with-keras): Hyperparameter tune a Keras model using HyperDrive, Azure ML's hyperparameter tuning functionality.
+* [train-and-deploy-to-aci](./vignettes/train-and-deploy-to-aci): Train a caret model and deploy as a web service to Azure Container Instances (ACI).
+* [train-with-tensorflow](./vignettes/train-with-tensorflow): Train a deep learning TensorFlow model with Azure ML.
+
+Find more information on the [official documentation site for Azure Machine Learning SDK for R](https://azure.github.io/azureml-sdk-for-r/).
+
+
+### Troubleshooting
+
+- If the following error occurs when submitting an experiment using RStudio:
+   ```R
+    Error in py_call_impl(callable, dots$args, dots$keywords) : 
+     PermissionError: [Errno 13] Permission denied
+   ```
+  Move the files for your project into a subdirectory and reset the working directory to that directory before re-submitting.
+
+  In order to submit an experiment, the Azure ML SDK must create a .zip file of the project directory to send to the service. However,
+  the SDK does not have permission to write into the .Rproj.user subdirectory that is automatically created during an RStudio
+  session. For this reason, the recommended best practice is to isolate project files into their own directory.
diff --git a/how-to-use-azureml/azureml-sdk-for-r/samples/README.md b/how-to-use-azureml/azureml-sdk-for-r/samples/README.md
@@ -0,0 +1,11 @@
+## Azure Machine Learning samples
+These samples are short code examples for using Azure Machine Learning SDK for R. If you are new to the R SDK, we recommend that you first take a look at the more detailed end-to-end [vignettes](../vignettes).
+
+Before running a sample in RStudio, set the working directory to the folder that contains the sample script in RStudio using `setwd(dirname)` or Session -> Set Working Directory -> To Source File Location. Each vignette assumes that the data and scripts are in the current working directory.
+
+1. [train-on-amlcompute](training/train-on-amlcompute): Train a model on a remote AmlCompute cluster.
+2. [train-on-local](training/train-on-local): Train a model locally with Docker.
+2. [deploy-to-aci](deployment/deploy-to-aci): Deploy a model as a web service to Azure Container Instances (ACI).
+3. [deploy-to-local](deployment/deploy-to-local): Deploy a model as a web service locally.
+
+> Before you run these samples, make sure you have an Azure Machine Learning workspace. You can follow the [configuration vignette](../vignettes/configuration.Rmd) to set up a workspace. (You do not need to do this if you are running these examples on an Azure Machine Learning compute instance).
diff --git a/how-to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/deploy-to-aci.R b/how-to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/deploy-to-aci.R
@@ -0,0 +1,59 @@
+# Copyright(c) Microsoft Corporation.
+# Licensed under the MIT license.
+
+library(azuremlsdk)
+library(jsonlite)
+
+ws <- load_workspace_from_config()
+
+# Register the model
+model <- register_model(ws, model_path = "project_files/model.rds",
+                        model_name = "model.rds")
+
+# Create environment
+r_env <- r_environment(name = "r_env")
+
+# Create inference config
+inference_config <- inference_config(
+  entry_script = "score.R",
+  source_directory = "project_files",
+  environment = r_env)
+
+# Create ACI deployment config
+deployment_config <- aci_webservice_deployment_config(cpu_cores = 1,
+                                                      memory_gb = 1)
+
+# Deploy the web service
+service <- deploy_model(ws,
+                        'rservice',
+                        list(model),
+                        inference_config,
+                        deployment_config)
+wait_for_deployment(service, show_output = TRUE)
+
+# If you encounter any issue in deploying the webservice, please visit
+# https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-troubleshoot-deployment
+
+# Inferencing
+# versicolor
+plant <- data.frame(Sepal.Length = 6.4,
+                    Sepal.Width = 2.8,
+                    Petal.Length = 4.6,
+                    Petal.Width = 1.8)
+# setosa
+plant <- data.frame(Sepal.Length = 5.1,
+                    Sepal.Width = 3.5,
+                    Petal.Length = 1.4,
+                    Petal.Width = 0.2)
+# virginica
+plant <- data.frame(Sepal.Length = 6.7,
+                    Sepal.Width = 3.3,
+                    Petal.Length = 5.2,
+                    Petal.Width = 2.3)
+
+# Test the web service
+predicted_val <- invoke_webservice(service, toJSON(plant))
+predicted_val
+
+# Delete the web service
+delete_webservice(service)
diff --git a/...to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/project_files/model.rds b/...to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/project_files/model.rds
diff --git a/how-to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/project_files/score.R b/how-to-use-azureml/azureml-sdk-for-r/samples/deployment/deploy-to-aci/project_files/score.R
@@ -0,0 +1,17 @@
+# Copyright(c) Microsoft Corporation.
+# Licensed under the MIT license.
+
+library(jsonlite)
+
+init <- function() {
+    model_path <- Sys.getenv("AZUREML_MODEL_DIR")
+    model <- readRDS(file.path(model_path, "model.rds"))
+    message("model is loaded")
+
+    function(data) {
+        plant <- as.data.frame(fromJSON(data))
+        prediction <- predict(model, plant)
+        result <- as.character(prediction)
+        toJSON(result)
+    }
+}