RandySheriffH
d3b684cd9e
Drop nuphar ( #11555 )
...
* drop nuphar code and configs
* refactor test case
* format python
* remove nuphar from training test
* remove commented nuphar logics
* restore llvm setting
* drop nuphar ci
* fix compile err
* fix compile err
Co-authored-by: Randy Shuai <rashuai@microsoft.com>
2022-09-07 15:11:18 -07:00
Jian Chen
acc8bdc6c5
Splitting quantize_tensor and quantize_input ( #12873 )
...
* Splitting quantize_tensor and quantize_input
* Reformat code
* Reformat code
* Update is_input_a_weight to is_input_a_initializer
2022-09-07 18:05:42 -04:00
Sheil Kumar
535b0835f2
User/sheilk/dft fixes ( #12862 )
...
* DirectML DFT Tests and Fixes
* Dynamicaly allocate temporaries using the allocator...
* Allocate during compute
* wrong dims
* CR feedback
2022-09-07 13:21:56 -07:00
sophies927
f63bd0765d
New GitHub templates ( #12777 )
...
* Create 01-build.yml
* Create 02-documentation.yml
* Create 03-mobile.yml
* Create 04-web.yml
* Create 05-performance.yml
* Create 06-training.yml
* Create 07-feature_request.yml
* Create 08-general.yml
* Create config.yml
* Delete bug-performance-issue.md
* Delete feature_request.md
* Create labeler.yml
* Create labeler.yml
* Update Performance template to make model info optional.
* Update feature request description placeholder
2022-09-07 11:59:56 -07:00
Hariharan Seshadri
ad69aac491
Introduce ordered quantization ops for the CUDA EP [1/n] ( #12582 )
...
Initial core small set for the ordered quantization ops for cuda EP.
2022-09-07 11:58:15 -07:00
petermcaughan
69f7cc6494
Add pybind support for all memory config options in OrtArenaCfg ( #12658 )
...
* Add support for initial_growth_chunk_size_bytes setting in OrtArenaCfg pybind
* Add overloaded constructor for KVP, UT still in progress
* Fix class member access in pybind, fix unit test
* Resolve linter warnings
* Improve formatting
* Simplify UT
* Fix linter formatting
Co-authored-by: Peter Mcaughan <petermca@microsoft.com>
2022-09-07 11:15:00 -07:00
Chen Fu
8004db4bf1
fix python import sequence warning ( #12864 )
...
fix python import sequence warning
2022-09-07 09:53:39 -07:00
Xavier Dupré
400195a10a
raise an exception when TreeEnsemble request a feature out of boundaries ( #12859 )
...
* Catch a potential error when the number of featues is low than the features referenced in TreeEnsemble
* add unit test
* remove extra spaces
2022-09-07 10:05:32 +02:00
Guenther Schmuelling
f856be162e
fix xnnpack wasm build ( #12845 )
2022-09-06 19:20:07 -07:00
Jan Tilly
437409c343
Add DONT_VECTORIZE flag to cmake ( #12169 )
...
Add DONT_VECTORIZE flag.
2022-09-07 12:14:14 +10:00
Scott McKay
706e03c63d
Add azp run helper ( #12832 )
...
* Add helper to add azp run comments to a PR.
2022-09-07 11:48:31 +10:00
Yi Zhang
c571b99336
Refactor setup_test_data ( #12818 )
...
* refactory setup_test_data
* mv setup test data to test stage
* model link for C# test
* add comment
2022-09-07 08:33:27 +08:00
Yulong Wang
726251609a
increase max memory to 4G for wasm ( #12798 )
2022-09-06 17:07:13 -07:00
Tianlei Wu
d19955fd89
fix transformers script issues ( #12802 )
...
Fix a few obvious issues:
(1) bert_perf_test.py create session without provider in line 65.
(2) compare_bert_results.py miss a parameter in create_session in line 37
(3) onnx_exporter.py returns value mismatch in lines 667, 690.
(4) remove some imports not used in the scripts.
(5) fusion_utils need not print "Removed 0 cast nodes" or "Removed 0 Identity nodes"...
(6) update requirements for numpy version since gpt2 parity tool use equal_nan in numpy v1.19+
2022-09-06 16:15:16 -07:00
Xavier Dupré
54360c88d2
Disable two warnings raised by tensorboard on Visual Studio ( #12773 )
2022-09-06 20:42:52 +02:00
Chen Fu
9ad5b95e4f
Fix math domain error with log10 ( #12841 )
...
fix math domain error with log10
2022-09-06 08:54:41 -07:00
Cheng
8cedafe250
[xnnpack] Have Initializer in Mobile related EPs in Minimal_build and creating EP specific dynamic-schema ( #12555 )
...
* Remove the dependence of Qlinearsoftmax schema
* refactor initializerview && create shared schema
* Dynamic Create EP specific schema
* Have Initializer in minimal_build
* address comments
* remove CancelFuseSubGraph
2022-09-06 14:32:15 +08:00
Scott McKay
ac4f1bf960
Update max opset for NNAPI and CoreML. ( #12831 )
...
Update max opset for NNAPI and CoreML. Changes in opsets 16 and 17 don't require any updates.
2022-09-05 09:37:14 +10:00
Baiju Meswani
9e47eb68e0
Remove unused orttraining amd dockerfiles and scripts ( #12707 )
2022-09-02 18:43:21 -07:00
Cheng
76d17b0f48
Add java API for xnnpack ( #12788 )
...
* Add java API for xnnpack
* provider option support
* a more general interface for creating EP
2022-09-03 08:29:40 +08:00
Baiju Meswani
295bd26980
Remove orttraining-distributed CI pipeline ( #12738 )
2022-09-02 14:34:26 -07:00
ashbhandare
27dde0b51f
Csharp bindings for on-device training APIs ( #12404 )
2022-09-02 13:13:48 -07:00
Jian Chen
2fe919c3ad
Adding Split Fusion ( #12732 )
...
* Adding Split Fusion
* Make changes to comments
* Format files and change typo
* Format files and change typo
* Format files and change typo
* Format files and change typo
* Format file
* Format files
* Format files
* Format files
* Format files
2022-09-02 14:17:10 -04:00
Baiju Meswani
56bae3b196
Use InplaceClipGradNorm for offline processing for on-device training ( #12603 )
2022-09-02 07:47:17 -07:00
Cassie Breviu
98b2b7f5bb
Update csharp documentation ( #12830 )
2022-09-01 22:14:03 -07:00
sophies927
548938fb97
Update stale.yml ( #12813 )
...
* Update stale.yml
Change the number of days of inactivity before an issue becomes stale from 60 to 5 and the number of days of inactivity before a stale issue is closed from 7 to 5. Update the exempt labels based on the redefined set of GH labels.
* Implement stale.yml feedback.
2022-09-01 20:50:46 -07:00
Changming Sun
ca5af24765
Update Sdl.ruleset to remove C26812 from the rules ( #12695 )
2022-09-01 20:05:20 -07:00
Hariharan Seshadri
931c8b0147
Resolve GH issue 12706 ( #12815 )
2022-09-01 18:30:57 -07:00
Justin Chu
6fe712b587
Create codeql.yml to replace LGTM ( #12790 )
...
**Description**: Create codeql.yml to replace LGTM
**Motivation and Context**
LGTM.com is shutting down and moving to github code scanning. This PR enables github code scanning.
cpp and c# support will be added in a separate pr.
2022-09-01 16:37:43 -07:00
ashbhandare
349469c381
Enable way to extract all parameters to and from a contiguous buffer. ( #12674 )
...
* implementation
* review comments
* review comment
* lint error
2022-09-01 15:23:30 -07:00
Hariharan Seshadri
52ce6a90b4
Props file cleanup ( #12782 )
2022-09-01 11:05:46 -07:00
George Nash
0125e15281
Fix include order build failure training build ( #12425 )
...
Signed-off-by: George Nash <george.nash@intel.com>
2022-09-01 10:48:40 -07:00
Sheil Kumar
e3b501125d
DFT on DirectML ( #12710 )
...
* DFT on DirectML
* feedback
* fix misc build issues
* fixes
* fix constant cpu inputs and optional tensors for external operators
* disable dft tests on 'pure' dml
2022-09-01 08:31:14 -07:00
Cheng
5dd9afe75a
python lint ( #12825 )
2022-09-01 22:38:25 +08:00
PeixuanZuo
adbc0757ad
[UPDATE] update ROCm ci pipeline to ROCm5.2.3 ( #12799 )
...
* [Update] update to rocm5.2.3
* [Fix] cmake version
* [Fix] disbale ortmodule tests
* [revert] revert performance number
2022-09-01 10:32:24 +08:00
Vincent Wang
262a597e2a
[CUDA] BiasSoftmax and Dropout Fusion ( #12667 )
...
* bias softmax dropout fusion
* fix rocm build
* move some files
2022-09-01 10:01:44 +08:00
Tianlei Wu
7e1e0f5c9e
fix build error for generation debug code ( #12801 )
...
fix dumping code
2022-08-31 13:53:01 -07:00
Justin Chu
a48b115540
Remove reference to the deprecated variable in torch.onnx.symbolic_helper ( #12452 )
...
**Description**: Remove reference to the deprecated variable in `torch.onnx.symbolic_helper` pytorch/pytorch#81953
- Removed unused imports
- Changed BANNED_AUTOGRAD_FUNCTION_NAMES to a frozenset
**Motivation and Context**
The cast_pytorch_to_onnx variable is deprecated and removed in `torch.onnx.symbolic_helper`. Since there is still a need for converting scalar types to onnx type, I copied the mapping to `_CAST_PYTORCH_TO_ONNX` in the module.
2022-08-31 11:55:56 -07:00
Dmitri Smirnov
f59a44e24f
Build VS 2022 no Abseil adjustment ( #12780 )
...
Build VS 2022 no Abseil
2022-08-31 11:47:43 -07:00
Nat Kershaw (MSFT)
cc94ba7740
Update C# API docs workflow ( #12796 )
2022-08-31 09:41:01 -07:00
Yi-Hong Lyu
6b148e15e1
Disable QGEMM, s8 A, s8 B (Packed) bench for AMD64 ( #12765 )
...
Besides, rename QGEMM tests for clarification.
2022-08-30 16:47:36 -07:00
Baiju Meswani
a52543ecd8
Generate windows training package ( #12789 )
2022-08-30 16:35:50 -07:00
Yulong Wang
82a28cc2c3
upgrade emsdk to 3.1.19 ( #12690 )
...
* upgrade emsdk to 3.1.19
* fix build break
* ignore '-Wunused-but-set-variable' in eigen
* add malloc and free in exported functions
* EXPORTED_FUNCTIONS
2022-08-30 13:42:45 -07:00
Yulong Wang
1a402a3f25
replace 'master' branch ref to 'main' for onnx repo ( #12678 )
2022-08-30 13:41:42 -07:00
Ye Wang
9aefcc251f
fix some prefast warnings ( #12730 )
...
fix warnings
2022-08-30 12:52:59 -07:00
cloudhan
9680ffd842
Fix rocm build caused by #12699 ( #12787 )
2022-08-30 20:26:16 +08:00
Yi Zhang
b4f6dad7c9
increase timeout limit of mac silicon package workflow ( #12784 )
...
increase timeout
2022-08-30 13:57:01 +08:00
cloudhan
9907b59a1e
Change cuda and rocm error checking helpers to return Status ( #12699 )
...
* CudaCall returns Status in non-throw and void in throw
* RocmCall returns Status in non-throw and void in throw
2022-08-30 13:18:47 +08:00
pengwa
a0c25e5c2f
Fix segment fault for alltoall ( #12701 )
...
* fix segment fault
* formatting
2022-08-30 11:27:14 +08:00
PeixuanZuo
19ca2a0089
[ADD] python package pipeline for ROCm5.2.3 ( #12770 )
...
* [TEST] test rocm5.2.3
[TEST] rm torchversion
[Update]sort
Co-authored-by: Ubuntu <peixuanzuo@peixuanzuomi200vm.zvflicr54joexhdgnhvmxrxygg.phxx.internal.cloudapp.net>
2022-08-30 11:05:59 +08:00
