…NNX_ATEN_FALLBACK when BUILD_CAFFE2 is ON (pytorch#67460)

The use of ATEN as a fallback operator during ONNX conversion is important for increasing operator coverage or even provide more efficient implementations over some ONNX ops.

Currently this feature is available through `OperatorExportTypes.ONNX_ATEN_FALLBACK`,
but it also performs changes to the graph that are runnable by Caffe2, only.

This PR introduces restricts caffe2-specific graph transformations for `ONNX_ATEN_FALLBACK`
operator export type for when pytorch is built with caffe2 support (aka BUILD_CAFFE2=1 during build)

The first version of this PR introduced a new operator export type `ONNX_ATEN__STRICT_FALLBACK`,
which essentially is the same as `ONNX_ATEN_FALLBACK` but without caffe2 transformations.
It was preferred to not introduce a new operator export type, but to refine the existing aten fallback one

## BC-breaking note
### The global constant `torch.onnx.PYTORCH_ONNX_CAFFE2_BUNDLE` is removed in favor of
a less visible `torch.onnx._CAFFE2_ATEN_FALLBACK`.
`PYTORCH_ONNX_CAFFE2_BUNDLE` is really a dead code flag always set to False.
One alternative would be fixing it, but pytorch#66658 disables Caffe2 build by default.
Making a Caffe2 feature a private one seems to make more sense for future deprecation.

### The method `torch.onnx.export` now defaults to ONNX when `operator_export_type` is not specified.
Previously `torch.onnx.export's operator_export_type` intended to default to `ONNX_ATEN_FALLBACK` when `PYTORCH_ONNX_CAFFE2_BUNDLE` was set, but it would never happen as `PYTORCH_ONNX_CAFFE2_BUNDLE` is always undefined

ghstack-source-id: 8ae8ac0
Pull Request resolved: pytorch#68490

* Allows implementing symbolic functions for domains other than `aten`, for example `prim`, in symbolic_opset#.py.
* Allows symbolic function to access extra context if needed, through `SymbolicFunctionState`.
  * Particularly, the `prim::PythonOp` special case can access node without the need of passing node through inputs. Updates will be made downstreams, and in a follow-up PR we will remove the previous workaround in exporter.
* `prim::Loop`, `prim::If`, etc are now moved outside of `_run_symbolic_function` from utils.py, and to symbolic_opset9.py.

Motivation for this change:
- Better maintainability and reducing complexity. Easier to add symbolic for operators, both simple and complex ones (that need additional context), without the former needing to know the existence of the latter.
- The design idea was long outdated. prim ops are no longer rare special cases, and they shouldn't all be handled inside `_run_symbolic_function`. As a result this function becomes too clumsy. There were also prim ops symbolic added in symbolic_opset#.py with signature `prim_[opname]`, creating separation and confusion.

Co-authored-by: BowenBao <bowbaomicrosoft.com>

ghstack-source-id: 577a7e6
Pull Request resolved: pytorch#68491

* Initial commit

* Fix flake issue

* Add test tags

ghstack-source-id: b277f18
Pull Request resolved: pytorch#68492

Fixes pytorch#66786.

`index_select` only supports `index` of 1-D tensor. `ONNX::Gather` allows `index` to have rank `q`. Abort constant folding `ONNX::Gather` if `index` rank is larger than 1.

Co-authored-by: BowenBao <bowbaomicrosoft.com>

ghstack-source-id: 9b53061
Pull Request resolved: pytorch#68493

Co-authored-by: Gary Miguel <garymiguelmicrosoft.com>

ghstack-source-id: a662f60
Pull Request resolved: pytorch#69544

The arg is not used and was previously deprecated.

Also remove torch.onnx._export_to_pretty_string. It's redundant with the
public version.

ghstack-source-id: 0d451e0
Pull Request resolved: pytorch#69546

ScriptModule export introduces duplicated ONNX initializers for shared weights, unnecessarily increases ONNX model size. This PR de-duplicates ONNX initializers for model exported in eval mode, by checking if the underlying tensors share the same `data_ptr`, `strides` and `sizes`.

Co-authored-by: BowenBao <bowbaomicrosoft.com>

ghstack-source-id: d17dfa4
Pull Request resolved: pytorch#69547

* Add Concat to Scalar type analysis pass

By using scalar type analysis for Concat, the exported model can do
automatic type promotion for Concat nodes, including mixed fp16 and fp32
inputs, for example.

Unit tests based on the original PR pytorch#24378

* Fix UTs

ghstack-source-id: 4e796b1
Pull Request resolved: pytorch#69548

[ONNX] minor clarifications of docstrings

1. Make description of ONNX_ATEN_FALLBACK more accurate (after pytorch#67460).
2. Specify minimum and maximum values for opset_version. This is pretty
   important information and we should make users dig through source
   code to find it.

ghstack-source-id: 6572ba2
Pull Request resolved: pytorch#69549

Fix the wiki URL.

Also minor reorganization in onnx.rst.

ghstack-source-id: e59c7f1
Pull Request resolved: pytorch#69550

[ONNX] restore documentation of public functions (pytorch#69623)

The build-docs check requires all public functions to be documented.
These should really not be public, but we'll fix that later.'

This PR adds a new attribute overload_name to the Aten node so that third party applications can implement calls to libtorch without using PyTorch source code.

This is necessary because torch's torch::jit::findOperatorFor(fullname) requires a full name, including operator and overload names.

ATen op was originally created for Caffe2, which leveraged the availability of the pytorch yaml files to create calls to the aten oeprators directly, not relying on torch::jit::findOperatorFor

The first part of the PR refactors all symbolics that create Aten ops, so that there is a single helper for this operator. Next all symbolics are updated to pass in the relevant overload name, if empty string is not applicable

Extend shape inference support for `Expand`, when value of argument `shape` is unknown. Infer the rank of the output of `Expand`, and set shape to dynamic, if shape of argument `shape` is known.

Without this, shape inference aborts, and falls back to the static shape provided by tracer, which is incorrect in many cases.

Co-authored-by: BowenBao <bowbao@microsoft.com>

…ear (pytorch#69232)

Co-authored-by: David Fan <jiafa@microsoft.com>

* Add module name as pythonOp attr

* Move to trace_post_record

* Add tests

* Code compactness

Enable tests that are fixed by ORT 1.10