-
Notifications
You must be signed in to change notification settings - Fork 3.9k
Logistics for ONNX Release 1.11.0
liqun Fu edited this page Feb 15, 2022
·
17 revisions
Release Manager: Liqun Fu
Target Release date: Week of 01/31/2022
- 01/26 - Create v1.11.0 release wiki with release schedule. (Done)
- 01/26 - Document all [1.11.0 planned work items for the release in this wiki. (Done)
- 02/03 - Code freeze. All PRs must be validated and merged by this date. (Done)
- 02/07 - Validate all new ops with ORT. (Done) Flush CI/CD pipeline (what is this?)
- 02/07 - Resolve pipeline failures, if any, and integrate any critical late PRs. (Done)
- 02/07 - Cut the release branch. (Done)
- 02/08 - Create test packages. See onnx v1.11.0rc
- 02/08 - Request partner validation of test packages. (Done)
- 02/14 - Create second test packages. See onnx v1.11.0rc
- 02/14 - Request partner validation of the second test packages. (Done)
- 02/16 - Complete ORT/partner/converters/community validation and approvals. (In Progess)
- 01/16 - ONNX v1.11.0 release and announcements. (In Progess)
all new and updated operators are to be validated by https://github.com/microsoft/onnxruntime/pull/10441
| Op Name | Description | Validation status | ONNX PR |
|---|---|---|---|
GridSample |
Add new operator | passed | #3557 |
Identity |
Add optional type support | passed | #3607 |
If |
Add optional data type support for output | passed | #3607 |
LeakyRelu |
Add bfloat16 type support | NA | #3960 |
Loop |
Add optional data type support for initial value and output | passed | #3607 |
PRelu |
Add bfloat16 type support | NA | #3960 |
RoiAlign |
add an attribute coordinate_transformation_mode, correct the default behavior | passed | #3625 |
Scan |
Add bfloat16 type support for output | NA | #3960 |
ScatterElements |
Add reduction attribute | passed | #3484 |
ScatterND |
Add reduction attribute | passed | #3484 |
Where |
Extend Where op to permit bfloat16 types | NA | #3738 |
GreaterOrEqual |
Add bfloat16 type support | NA | #3960 |
LessOrEqual |
Add bfloat16 type support | NA | #3960 |
| Op Name | Description | Validation status | ONNX PR |
|---|---|---|---|
TreeEnsembleClassifier |
support tensor as attributes | NA | #3897 |
TreeEnsembleRegressor |
support tensor as attributes | NA | #3897 |
| Description | PR | Status | Notes |
|---|---|---|---|
| Add compose utils | #3820 | Merged | jantonguirao |
| Description | PR | Status | Notes |
|---|---|---|---|
| Extend optional type inference | #3756 | Merged | neginraoof |
| Make shape inference handle MapProto | #3772 | Merged | jcwchen |
| Improve rank inference for Expand op | #3807 | Merged | TomWildenhain-Microsoft |
| Enhance shape inference: ParseData/Transpose/QuantizeLinear | #3806 | Merged | jcwchen |
| Honor existing dim_param in shape inference | #3896 | Merged | shinh |
| Shapeinf for functions | #3722 | Merged | askhade |
| Use symbolic input for shape inference of ConstantOfShape | #3784 | Merged | xuzijian629 |
| Description | PR | Status | Notes |
|---|---|---|---|
| Use MSVC Runtime as dll for official ONNX Windows release | #3644 | Merged | jcwchen |
| Simplify common version converter adapter design patterns | #3761 | Merged | gramalingam |
| Use scalar for OneHot's depth to prevent confusion | #3774 | Merged | jcwchen |
| Correct wrong subgraph test example for If operator | #3798 | Merged | jcwchen |
| [Dup] Add SpaceToDepth test cases | #3786 | Merged | jcwchen |
| Fix error in Pad op convert | #3778 | Merged | liuyu21 |
| Fix some examples for ArgMax | #3851 | Merged | yihonglyu |
| Shape inference should not propagate missing optional outputs | #3815 | Merged | jcwchen |
| Check negative index for attributes of Slice-1 | #3810 | Merged | jcwchen |
| Cleanup type cast related warnings | #3801 | Merged | gongsu832 |
| replace whitelist by safelist | #3900 | Merged | AlexandreEichenberger |
| Fix weekly/Linux CI failures: correct skip list and remove old numpy related code | #3916 | Merged | jcwchen |
| Fix old ConvTranspose shape inference and softmax upgrader | #3893 | Merged | matteosal |
| Fix Linux i686 Release CI failure due to the latest NumPy | #3918 | Merged | jcwchen |
| Simplify function definition of context-dependent functions | #3882 | Merged | gramalingam |
| Migration to using main branch | #3925 | Merged | jcwchen |
| Append dim even both dim value and param are not set | #3828 | Merged | xuzijian629 |
| Bump to 10.15 in AzurePipeline because 10.14 was deprecated | #3941 | Merged | jcwchen |
| six: remove all references | #3926 | Merged | stillmatic |
| for issue 3849 to confirm that type check is performed during checker | #3902 | Merged | liqunfu |
| Remove testing ort-nightly for Mac Python 3.6 due to unsupported ort-nightly | #3953 | Merged | jcwchen |
| mypy: update to 0.760 and remove vendored protobuf stubs | #3939 | Merged | stillmatic |
| Upgrade Windows version in AzurePipeline since 2017 was dep | #3957 | Merged | jcwchen |
| Version converter for Softmax should not produce empty shape | #3861 | Merged | jcwchen |
| Fix Cppcheck warning about memset on NULL backend_ids | #3970 | Merged | cjvolzka |
| Bugfix extractor misses local functions | #3954 | Merged | liqunfu |
| Add bfloat16 type to a few ops missing it | #3960 | Merged | gramalingam |
| Description | PR | Status | Notes |
|---|---|---|---|
| ONNX Hub Docs | #3712 | Merged | mhamilton723 |
| Clarify definition of a tensor in IR docs | #3792 | Merged | TomWildenhain-Microsoft |
| Document that Where supports multidirectional broadcasting | #3827 | Merged | garymm |
| Sync build documentation in CONTRIBUTING.md | #3859 | Merged | jcwchen |
| [CI][Doc] Add CI Pipelines doc/node tests verification | #3780 | Merged | jcwchen |
| Remind release manager to remove old onnx-weekly packages after release | #3923 | Merged | jcwchen |
| Fix the bug of shape in docs | #3927 | Merged | namasikanam |
| Clean up README | #3961 | Merged | garymm |
| Remove documentation about Python 2 | #3963 | Merged | garymm |
The following testing can be added into release pipelines after producing the wheel to let release manager be aware.
- Test with the latest ORT with onnx/test/test_with_ort.py
- Test with different versions of dependencies (protobuf, numpy). Take protobuf as an example: test with 3.11.3 and the latest one.
| Description | Owner | Status |
|---|---|---|
| Run local sanity tests | Liqun Fu | In progress |
| Cut v1.11 release branch | Liqun FU | done |
| Validate PyPI test packages | Liqun Fu, Ashwini Khade, Jacky Chen | In progress |
| Partner validation of test packages | Various | sent a testpipy rc link to partners for validation |
| Validate Opset 16 with ORT | Liqun Fu | done |
| Create release summary | Liqun Fu | In progress |
| Publish ONNX v1.11.0 | Liqun Fu |
- pytorch
- onnx-tensorflow
- tensorflow-onnx
- sklearn-onnx
- onnxmltools
- keras-onnx
- onnx-tensort
- onnx-coreml
Thanks to these individuals for their contributions in this release since last 1.10.0 release. (Contributor list obtained with: https://github.com/onnx/onnx/graphs/contributors?from=2021-07-30&to=2022-02-08&type=c): @jcwchen, @gramalingam, @garymm, @mhamilton723, @TomWildenhain-Microsoft, @neginraoof, @xuzijian629, @liqunfu, @gwang-msft, @chudegao, @AlexandreEichenberger, @rajeevsrao, @matteosal, @stillmatic, @askhade, @liuyu21, @jantonguirao, @shinh, @kevinch-nv, @shubhambhokare1, @hwangdeyu, @jiafatom, @postrational, @snnn, @jackwish