Commit graph

6560 commits

Author SHA1 Message Date
Adrian Tsai
ae08f9666d
Fix type constraints in registration of DequantizeLinear (#10986) 2022-03-23 17:05:12 -07:00
Sheil Kumar
938f3857a5
Set the default for the STFT onesided attribute to 1, which tests expect (#10984)
Co-authored-by: Sheil Kumar <sheilk@microsoft.com>
2022-03-23 14:20:54 -07:00
Chandru Ramakrishnan
07201726ed
Fixed macros for graph transformer registration. (#10983) 2022-03-23 14:55:17 -04:00
Olivia Jain
de384805cd
Custom parameters (#10964)
* get inputs independently for trtexec

* track one process only

* remove engine and profile files

* change time to commit time

* add runtime option for io binding

* move to commit date

* fixes

* add option for graph optimization

* cleanup docker script

* note second time creation

* allow for parameters to be configured from pipeline at runtime

* uncomment

* include optional arguments at runtime

* post second session creation

* update cmake version

* Revert "update cmake version"

This reverts commit 09a1364eae68610724c8e90eeea777b7ee03f74b.

* Move data format import
2022-03-23 09:47:24 -07:00
Jeff Daily
9a3be9b46a
use #include <hiprand/hiprand.h>, not deprecated #include <hiprand.h> (#10966) 2022-03-23 08:56:45 -07:00
Yi Zhang
0efbe92296
fix coverage report error in master build (#10969)
* fix error in master

* check NNAPI_EP_MASTER

* Revert "check NNAPI_EP_MASTER"

This reverts commit 59c9043b7c9bbcb4b495d2dd121ef6d4271be408.

* rm coverage in PR build
2022-03-23 16:00:57 +08:00
raviskolli
480c793125
Update training packages to Pytorch 1.11.0 (#10851)
* Update ortmodule training packages to Pytorch 1.11.0

Co-authored-by: Harshitha Venkata <havenka@microsoft.com>
Co-authored-by: Baiju Meswani <bmeswani@microsoft.com>
2022-03-22 16:45:51 -07:00
Baiju Meswani
565318ce86
Support ORT WASM compilation with the training flag (#10973)
* Add training support for ORT web assembly compilation

* Use wrapper for eigen includes in training
2022-03-22 16:13:35 -07:00
Scott McKay
b28e5064f3
Ignore DequantizeLinear nodes in CommonSubexpressionElimination optimizer (#10934)
* Ignore DequantizeLinear nodes in CommonSubexpressionElimination.

Coalescing DQ nodes results in QDQ node groups having overlaps, which the QDQ processing does not support.
2022-03-23 08:46:01 +10:00
Xavier Dupré
b88fb68fac
Adds missing numpy type when looking for the ort correspondance (#10943) 2022-03-22 14:44:48 -07:00
Yulong Wang
dce5d719c5
add build flag for emscripten settings (#10963)
* allows multiple '--cmake_extra_defines' flags

* fix flake8 error

* Add build flag for emscripten settings

* remove "emscripten_settings" in generate_build_tree()

* format code
2022-03-22 11:55:45 -07:00
Sheil Kumar
027565b3b2
Add multi-dim dft test, and fix complex idft (#10947)
* fix complex multi-dim dft

* Add multi-dim dft test, and fix complex idft

* remove incorrect inplace specification

* Add DFT tests

* update epsilon to 1000ths place

Co-authored-by: Sheil Kumar <sheilk@microsoft.com>
2022-03-22 10:08:12 -07:00
Yulong Wang
2da82fd0b9
allows multiple '--cmake_extra_defines' flags (#10953)
* allows multiple '--cmake_extra_defines' flags

* fix flake8 error
2022-03-21 19:10:47 -07:00
Sunghoon
6d19c295d0
use lf as eol for node package (#10965) 2022-03-21 15:50:03 -07:00
Sunghoon
b34d9f6867
[js/wasm] Add WebAssembly static library build into web CI pipeline (#10959)
* add webassembly static library build into ci

* add webassembly static library build into ci

* skip publishing on static lib

* fix type
2022-03-21 15:49:49 -07:00
Chandru Ramakrishnan
4a5b5328a4
Added support to Eager CodeGen for multiple in-place parameters. (#10945)
* Added support to CodeGen for multiple inplace output parameters.

* Updated output Tensor to references.
2022-03-21 13:10:22 -07:00
Leandro Gracia Gil
1cc2cfb7b8
Move #ifndef ORT_CXX_API_THROW to the no exceptions case. (#10937)
This is related to https://github.com/microsoft/onnxruntime/issues/10564
which introduced a fix in the wrong case where exceptions are enabled.
2022-03-21 11:12:56 -07:00
leqiao-1
a6ea278502
add python3.10 support (#10848)
* add python3.10 support

* upgrade numpy version in build pipeline

* add python 3.10 path

* upgrade torch version in build pipeline

* update docker run arguments

* change torch version

* fix typo

* fix permission issue

* change python version

* remove python3.10 for openvino build

* remove python 3.10 for openvino build
2022-03-21 09:46:02 +08:00
G. Ramalingam
8703d37517
Extend DropoutGrad function to support bfloat16 (#10662)
* Update DropoutGrad function to support bfloat16

* Eliminate dead comments

* Set opset version for testcase

Signed-off-by: Ganesan Ramalingam <grama@microsoft.com>

* Update to new builder

Signed-off-by: Ganesan Ramalingam <grama@microsoft.com>
2022-03-20 15:11:08 -07:00
Scott McKay
91722e2bc4
Fix typos (#10935) 2022-03-20 08:27:35 +10:00
Yi Zhang
c1e37e4ebf
Android CI Pipeline: Fix post coverage bug (#10949) 2022-03-19 11:17:08 -07:00
Ella Charlaix
fe6ab719f3
Fix a typo in quantization tools (#10940) 2022-03-18 21:03:16 -07:00
soundarthiaga
eabb14788a
[perf_metric] added inferences per second metric (#10921) 2022-03-18 21:01:11 -07:00
Yi Zhang
3897b93606
optimize Android CI (#10938) 2022-03-19 11:00:21 +08:00
Kotaro Yamamoto
2dea7dc27f
Skip python arena shrinkage test on ppc (#10901) 2022-03-18 19:31:21 -07:00
soundarthiaga
de06d95096
[parallel_inference] added support for parallel inference with timed duration perf test (#10922) 2022-03-18 19:05:28 -07:00
Scott McKay
5cbacec854
Maintain aspect ratio by doing resize + crop in image_to_pb tool (#10887) 2022-03-19 07:08:45 +10:00
ytaous
f058c59407
Performance: add io_binding support for bert benchmark util (#10907)
* io_binding support

* cover all test cases

* per comments

Co-authored-by: Ethan Tao <ettao@microsoft.com@orttrainingdev7.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net>
2022-03-18 10:33:30 -07:00
Changming Sun
c29d94383d
Update CODEOWNERS (#10932) 2022-03-18 09:37:58 -07:00
Ye Wang
ee05c591e5
Fix benchmark bugs and add Pytorch version control (#10928) 2022-03-18 09:24:19 -07:00
Guoyu Wang
6f844522c8
Follow up update for python API checking if vcruntime140_1.dll is available (#10927) (#10933) 2022-03-18 08:09:16 -07:00
Vincent Wang
8860fded02
Disable Some Einsum ORTModule Tests Due to Issue from PyTorch Exporter (#10906)
* disable some einsum tests due to pytorch issue

* disable tests on specific torch versions

* use skipif
2022-03-18 21:28:18 +08:00
Tiago Koji Castro Shibata
5ed2f4ad5f Remove Windows Store specific code 2022-03-17 23:38:14 -07:00
Changming Sun
a3eeb7b0f7 Update orttraining-linux-external-custom-ops.yml for Azure Pipelines 2022-03-17 23:36:53 -07:00
Ye Wang
78133434b5
Fix fp16 converter bugs[1/n] (#10882)
handle sequence type
2022-03-17 22:38:43 -07:00
zhangyaobit
5d4ff67c36
Support fusion options for benchmark.py (#10900)
* Support fusion options for benchmark.py

* Add fusion options for tf model export as well.

* Add command example and warning related to fusion options.
2022-03-17 20:57:43 -07:00
Guoyu Wang
b86d105153
[python API] Change raise import error when C:\Windows\System32\vcruntime140_1.dll is not found to warning (#10927)
* remove throw if C:\\Windows\\System32\\vcruntime140_1.dll cannot be found

* Add comments and update warning message

* adding back accidentally removed line

Co-authored-by: gwang0000 <62914304+gwang0000@users.noreply.github.com>
2022-03-17 18:56:43 -07:00
Sunghoon
740870a285
[js] Create npm packaging pipeline (#10886)
* create npm packaging pipeline

* fix indentations

* Update npm-packaging-pipeline.yml for Azure Pipelines

* Update npm-packaging-pipeline.yml for Azure Pipelines

* Update npm-packaging-pipeline.yml for Azure Pipelines

* react-native-ci as a template

* fix typos

* fix template paths

* add a depencendy

* change a stage name

* set different artifact name for each package

* fix typo

* Update npm-packaging-pipeline.yml for Azure Pipelines

Set a build Id for node npm package as a parameter

* Update npm-packaging-pipeline.yml for Azure Pipelines

Set a build Id for node npm package as a parameter

* Update npm-packaging-pipeline.yml for Azure Pipelines
2022-03-17 16:44:00 -07:00
Valery Chernov
766e6ac4fd
check TVM target for CPU (#10926)
Co-authored-by: Valery Chernov <valery.chernov@deelvin.com>
2022-03-17 15:51:24 -07:00
Yi-Hong Lyu
a46b00499a
Revert "Upsample support NHWC (#10554)" (#10917)
This reverts commit bd08f11a58.

Co-authored-by: Yufeng Li <liyufeng1987@gmail.com>
2022-03-17 13:16:29 -07:00
Sheil Kumar
810c18e809
fix complex multi-dim dft (#10896)
Co-authored-by: Sheil Kumar <sheilk@microsoft.com>
2022-03-17 12:45:51 -07:00
Edward Chen
07a71d5bf2
Fix handling of nodes inserted by NHWC transformer. (#10904) 2022-03-17 12:41:34 -07:00
cloudhan
e03b799b95
Make it clear in verbose log on why a kernel impl does not fit for a node (#10872) 2022-03-17 20:05:11 +08:00
Vincent Wang
6c0eff1ae4
optimize Transpose3DKernel (#10891) 2022-03-17 18:09:04 +08:00
PeixuanZuo
463fac67a3
[FIX] symbolic shape infer error with onnx-1.11.0 (#10674)
* [FIX] symbolic shape infer error with onnx-1.11.0

* [FIX] consider inputs name contains 'unk__'

* [TEST] enable gpt2 test

* [FIX] gpt2_megatron_opt.onnx graph
2022-03-17 13:47:02 +08:00
mindest
d7d7665023
restore random states after export_model (#10705)
* restore random states after export_model

* move get/set_random_states inside _export_model

* add comments for random state save/restore

* add unit test for random state check

* resolve comments

* fix error
2022-03-17 11:56:25 +08:00
Chi Lo
42d7112f03
Add multithreading test and put a lock on nvinfer1::createInferRuntime() for TRT EP (#10714)
* Add multithread unit test and put lock on library call

* update code

* remove debug code

* add comment

* add one session multi-threads inference

* Put lock for build engine all the time

* Update naming and comment

* remove unnecessary lock

* Revert "remove unnecessary lock"

This reverts commit 9c2317b1d2273dec0ebdeb52160bc757839e5edc.
2022-03-16 09:19:33 -07:00
Chi Lo
ce204d0744
Update to flatbuffers v2.0.0 (#10866) 2022-03-16 09:18:49 -07:00
Chun-Wei Chen
bac9c0eb50
skip optional related models from opset16 (#10840) (#10878) 2022-03-16 08:49:42 -07:00
PeixuanZuo
5763657715
[UPDATE] Add prefix in front of the file (#10884) 2022-03-16 21:05:18 +08:00