Merge pull request #106055 from lobrien/1669159-pytorch

PyTorch clarifications in TensorBoard & ONNX articles

Author: PRMerger17

articles/machine-learning/how-to-deploy-and-where.md

@@ -9,7 +9,7 @@ ms.topic: conceptual
 ms.author: jordane
 author: jpe316
 ms.reviewer: larryfr
-ms.date: 12/27/2019
+ms.date: 02/27/2020
 
 ms.custom: seoapril2019
 ---
@@ -893,6 +893,9 @@ service_name = 'onnx-mnist-service'
 service = Model.deploy(ws, service_name, [model])
 ```
 
+If you're using Pytorch, [
+Exporting models from PyTorch to ONNX](https://github.com/onnx/tutorials/blob/master/tutorials/PytorchOnnxExport.ipynb) has the details on conversion and limitations. 
+
 ### Scikit-learn models
 
 No code model deployment is supported for all built-in scikit-learn model types.

articles/machine-learning/how-to-monitor-tensorboard.md

@@ -8,7 +8,7 @@ ms.subservice: core
 ms.topic: conceptual
 author: maxluk
 ms.author: maxluk
-ms.date: 11/08/2019
+ms.date: 02/27/2020
 ---
 
 # Visualize experiment runs and metrics with TensorBoard and Azure Machine Learning
@@ -178,6 +178,10 @@ tb.start()
 tb.stop()
 ```
 
+>[!Note]
+ While this example used TensorFlow, TensorBoard can be used as easily with PyTorch or Chainer models. TensorFlow must be available on the machine running TensorBoard, but is not necessary on the machine doing PyTorch or Chainer computations. 
+
+
 <a name="export"></a>
 
 ## Option 2: Export history as log to view in TensorBoard