Commit Graph

6 Commits

Author SHA1 Message Date
Justin Chu
563c2719bf [ONNX] Refactor to remove inline imports - attempt 2 (#77448)
Re-land
- #77142

(diff: https://github.com/pytorch/pytorch/compare/c08b8f0..justinchuby:justinchu/remove-patch2)

Fixed:
- Delay import symbolic_opsets in the registry.

Tested locally with torchvision
Pull Request resolved: https://github.com/pytorch/pytorch/pull/77448
Approved by: https://github.com/garymm
2022-05-16 14:44:24 +00:00
PyTorch MergeBot
6b366dd3c1 Revert "[ONNX] Refactor to remove inline imports (#77142)"
This reverts commit c08b8f0967.

Reverted https://github.com/pytorch/pytorch/pull/77142 on behalf of https://github.com/malfet
2022-05-13 19:44:17 +00:00
Justin Chu
c08b8f0967 [ONNX] Refactor to remove inline imports (#77142)
Reduce circular dependencies

- Lift constants and flags from `symbolic_helper` to `_constants` and `_globals`
    - Standardized constant naming to make it consistant
- Make `utils` strictly dependent on `symbolic_helper`, removing inline imports from symbolic_helper
- Move side effects from `utils` to `_patch_torch`

Pull Request resolved: https://github.com/pytorch/pytorch/pull/77142
Approved by: https://github.com/garymm, https://github.com/BowenBao
2022-05-13 03:46:33 +00:00
Justin Chu
5dd1c67776 [ONNX] Format ONNX python with black
Format all onnx python code with black and isort with

```sh
isort torch/onnx/ test/onnx
black torch/onnx/ test/onnx
```

Updated lintrunner config to include these paths.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/76754
Approved by: https://github.com/suo, https://github.com/BowenBao
2022-05-05 00:19:22 +00:00
Edward Z. Yang
a11c1bbdd0 Run Black on all of tools/
Signed-off-by: Edward Z. Yang <ezyangfb.com>

Pull Request resolved: https://github.com/pytorch/pytorch/pull/76089

Approved by: https://github.com/albanD
2022-04-20 17:29:41 +00:00
Gary Miguel
ca374773b4 [ONNX] update default opset_version to 13 (#73898)
Summary:
And add a new tool to update it in the future, which follows the policy
of using "latest as of 18 months ago". This policy is meant to balance:
* recent enough to increase the odds of being able to successfully
  export
* old enough to increase the odds of exported model being runnable by
  different ONNX implementations

Related changes:

* test_models.py: explicitly fix opset_version to 9 rather than relying on default. Caffe2 doesn't support newer versions.
* symbolic_helper.py:
  * Remove a misleading comment
  * Remove unnecessary check in `_set_opset_version`
  * Use a range to define `_onnx_stable_opsets`
* test_pytorch_common.py:
  * Rename a variable from min -> max. I think it was a copy-paste error.
  * Make skip test messages more informative.
  * Remove unused `skipIfONNXShapeInference`. More on that below.
* test_pytorch_onnx_onnxruntime.py:
  * Make all the `TestCase` classes explicitly specify opset version.
  * Make `test_unsupported_pad` respect `opset_version` by using `run_test`
  * Unrelated simplification: make it obvious that all tests run with `onnx_shape_inference=True`. AFAICT this was already the case.
  * There was one test that was entirely disabled (test_tolist) because it was asking to be skipped whenever `onnx_shape_inference=True`, but it was always True. I changed the model being tested so as to preserve the intended test coverage but still have the test actually pass.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/73898

Reviewed By: msaroufim

Differential Revision: D35264615

Pulled By: malfet

fbshipit-source-id: cda8fbdffe4cc8210d8d96e659e3a9adf1b5f1d2
(cherry picked from commit b5e639e88828d34442282d0b50c977e610a2ba3a)
2022-04-07 00:02:31 +00:00