Add documentation for exporting model from PyTorch to different ONNX opset versions (#120)

* Add Doc For Export from PyTorch To ONNX

* Add batchnorm img

* Rename md file

* Update README file

* Update README file

* Update README file

Author: KsenijaS

README.md

@@ -21,7 +21,7 @@
 | [Menoh](https://github.com/pfnet-research/menoh) | [pfnet-research/menoh](https://github.com/pfnet-research/menoh) | n/a | [Importing](tutorials/OnnxMenohHaskellImport.ipynb) |
 | [ML.NET](https://github.com/dotnet/machinelearning/) | [built-in](https://www.nuget.org/packages/Microsoft.ML/) | [Exporting](https://github.com/dotnet/machinelearning/blob/master/test/Microsoft.ML.Tests/OnnxConversionTest.cs) | [Importing](https://github.com/dotnet/machinelearning/blob/master/test/Microsoft.ML.OnnxTransformerTest/OnnxTransformTests.cs) |
 | [Apache MXNet](http://mxnet.incubator.apache.org/) | part of mxnet package [docs](http://mxnet.incubator.apache.org/api/python/contrib/onnx.html) [github](https://github.com/apache/incubator-mxnet/tree/master/python/mxnet/contrib/onnx) | [Exporting](tutorials/MXNetONNXExport.ipynb) | [Importing](tutorials/OnnxMxnetImport.ipynb) |
-| [PyTorch](http://pytorch.org/) | [part of pytorch package](http://pytorch.org/docs/master/onnx.html) | [Exporting](tutorials/PytorchOnnxExport.ipynb), [Extending support](tutorials/PytorchAddExportSupport.md) | coming soon |
+| [PyTorch](http://pytorch.org/) | [part of pytorch package](http://pytorch.org/docs/master/onnx.html) | [Exporting](tutorials/PytorchOnnxExport.ipynb), [Exporting - Different ONNX Opsets](tutorials/ExportModelFromPyTorchToDifferentONNXOpsetVersions.md), [Extending support](tutorials/PytorchAddExportSupport.md) | coming soon |
 | [SciKit-Learn](http://scikit-learn.org/) | [onnx/sklearn-onnx](https://github.com/onnx/sklearn-onnx) | [Exporting](http://onnx.ai/sklearn-onnx/index.html) | n/a |
 | [TensorFlow](https://www.tensorflow.org/) | [onnx/onnx-tensorflow](https://github.com/onnx/onnx-tensorflow) and [onnx/tensorflow-onnx](https://github.com/onnx/tensorflow-onnx) | [Exporting - ONNX-Tensorflow](tutorials/OnnxTensorflowExport.ipynb)<br>[Exporting - Tensorflow-ONNX](https://github.com/onnx/tensorflow-onnx/blob/master/examples/call_coverter_via_python.py) | [Importing](tutorials/OnnxTensorflowImport.ipynb) [experimental] |
 | [TensorRT](https://developer.nvidia.com/tensorrt) | [onnx/onnx-tensorrt](https://github.com/onnx/onnx-tensorrt) | n/a | [Importing](https://github.com/onnx/onnx-tensorrt/blob/master/README.md) |

tutorials/ExportModelFromPyTorchToDifferentONNXOpsetVersions.md

@@ -0,0 +1,78 @@
+### Export model from PyTorch to ONNX:
+
+Default version for ONNX model is 9 when it is exported from PyTorch.
+
+```python
+import torch
+import torchvision
+
+# No CUDA
+dummy_input = torch.randn(10, 3, 224, 224)
+model = torchvision.models.resnet18(pretrained=True)
+
+# CUDA
+dummy_input = torch.randn(10, 3, 224, 224, device='cuda')
+model = torchvision.models.alexnet(pretrained=True).cuda()
+
+input_names = [ "actual_input_1" ] + [ "learned_%d" % i for i in range(16) ]
+output_names = [ "output1" ]
+
+torch.onnx.export(model, dummy_input, "resnet18.onnx", verbose=True, input_names=input_names, output_names=output_names)
+```
+
+### Version Conversion
+
+Version Conversion for BatchNormalization from opset 8 to 9:
+
+<img src="assets/batchnorm.png" />
+
+
+### Downgrade Version Conversion from 9 to 8:
+
+```python
+import onnx
+
+# Load the model
+model = onnx.load("path_to/resnet18.onnx")
+
+# Check that the IR is well formed
+onnx.checker.check_model(model)
+
+from onnx import version_converter
+
+# Convert to version 8
+converted_model = version_converter.convert_version(onnx_model, 8)
+
+# Save model
+onnx.save(converted_model, "path_to/resnet18_v8.onnx")
+```
+
+### Upgrade Version Conversion from 8 to 9
+
+```python
+# Convert to version 9
+converted_model9 = version_converter.convert_version(converted_model, 9)
+
+# Save model
+onnx.save(converted_model9, "path_to/resnet18_v9.onnx")
+```
+
+### Downgrade Version Conversion from 8 to 7
+
+```python
+# Convert to version 7
+converted_model7 = version_converter.convert_version(converted_model, 7)
+
+# Save model
+onnx.save(converted_model7, "path_to/resnet18_v7.onnx")
+```
+
+### Upgrade Version Conversion from 7 to 9
+
+```python
+# Convert to version 9
+converted_model79 = version_converter.convert_version(converted_model7, 9)
+
+# Save model
+onnx.save(converted_model79, "path_to/resnet18_v79.onnx")
+```