Update eager Readme.md (#12170)

orttraining/orttraining/eager/README.md

@@ -1,17 +1,50 @@
 # ONNX Runtime Eager Mode Support for PyTorch
 
-* Until the ONNX Runtime Eager Mode backend is merged in PyTorch upstream ([PR #58248](https://github.com/pytorch/pytorch/pull/58248)), build PyTorch from [abock/pytorch](https://github.com/abock/pytorch/tree/dev/abock/ort-backend) (branch: `dev/abock/ort-backend`):
+## Build and test
+1. sudo apt-get install python3-dev
+2. Upgrade CMAKE to > version 3.18.1.
+3. Create and activate a virtual environment:
+		a. python3 -m venv ~/venvs/onnxruntime
+		b. source ~/venvs/onnxruntime/bin/activate
+4. pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cpu
+5. pip3 install parameterized
+6. pip install packaging wheel
+7. Sudo apt-get install mpich libmpich-dev libcomp-11-dev (or use '--use_mpi=False' in the build command)
+8. Sudo apt-get update (optional if you think software is out of date and it might show out dated stuff to remove)
+9. Clone microsoft/onnxruntime
 
-  ```bash
-  $ git clone https://github.com/abock/pytorch -b dev/abock/ort-backend --recursive
-  $ cd pytorch
-  $ python setup.py develop --user
-  ```
+To build, go to the root of the onnxruntime repository and run:
+./build.sh --cmake_generator Ninja --config Debug --build_shared_lib --parallel --build_eager_mode --enable_training --enable_pybind  --build_wheel --skip_tests
 
-* If not build as noted above, ensure the eager-mode enabled PyTorch build is installed in the current environment.
+To run the eager tests:
+ PYTHONPATH=~/{onnxruntime repo}/build/Linux/Debug python3 ~/{onnxruntime repo}/orttraining/orttraining/eager/test/ort_ops.py
 
-* Build ONNX Runtime, making sure you enable training and python build together with eager mode:
+## Mapping aten cpp namespace functions to onnx ops
 
-  ```bash
-  ./build.sh --cmake_generator=Ninja --enable_training --enable_pybind --build_eager_mode
-  ```
+Useful links
+- [Onnx Op Schema](https://github.com/onnx/onnx/blob/main/docs/Operators.md)
+- [Aten native ops](https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/native/native_functions.yaml)
+
+For mapping existing aten ops to onnx files start in `orttraining/orttraining/eager/opgen/opgen/atenops.py.` This file
+drives the generator which the build runs and produces `orttraining/orttraining/eager/ort_aten.g.cpp`. Looking at the
+generated code will be helpful to understand the aten op signature and how onnx ops are invoked.
+
+
+In the simple case, add the aten op and the mapped onnx op to the `hand_implemented` list. As
+an example `"aten::t": Transpose("self"),` maps the aten transpose function `t` to the onnx op `Transpose` and maps the
+`self` input param of `t` to the first argument of `Transpose`.
+
+
+Often mapping ten to onnx is not one to one, so composing is supported. Example
+`"aten::zeros_like": ConstantOfShape(Shape("self")),`.
+
+
+In some case more data shaping is required, so only the signature should be created such as `"aten::equal": SignatureOnly(),`.
+The implementation is then added to `orttraining/orttraining/eager/ort_aten.cpp`
+
+Please add tests for all ops. Tests are defined in `orttraining/orttraining/eager/test/ort_ops.py`
+
+## Decisions worth noting
+- Resizing the output tensor. Aten supports resizing any out tensor but prints a warning this is depracated and support
+will end. With that in mind, we have decided to error in `resize_output` if the output tensor is not empty or already
+the right shape.