saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-06-02 23:39:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
3426 commits 2856 branches 57 tags 1.3 GiB
Commit graph

3426 commits

This branch
This branch
All branches
Author SHA1 Message Date
Tianlei Wu
3bbce69185
bump version to 1.5.1 (#5258) 2020-09-22 20:57:34 -07:00
Jeff Bloomfield
59e69bf35b
Handle missing initializers in allocation planner to fix crashes with DML provider (#5244) 
* Fix memory planning bug with DML EP

* Address PR comments

* Fix typo
 2020-09-22 19:37:07 -07:00
Ye Wang
898531f502
Fix reshape fusion crash (#5252) 
* fix reshape fusion crash

* handling start_node statelessly

* fix
 2020-09-22 15:04:13 -07:00
Guoyu Wang
e30530d9ea
Add java API for AddSessionConfigEntry (#5241) 
* Add session option config entry API for java

* Java format

* Add extra test verification

* Address PR comments

* Update comments

Co-authored-by: gwang0000 <62914304+gwang0000@users.noreply.github.com>
 2020-09-22 14:51:39 -07:00
KeDengMS
8dceebda0e
[Training/Python] Add option to enable symbolic shape inference (#5107) 
This change adds symbolic shape inference to ORT training which helps static memory planning for model like BART.
 2020-09-22 10:49:07 -07:00
edgchen1
14f250a4d0
Update BUILD.md training dependency info. (#5240) 
Update training dependency versions based on Dockerfile.training.
 2020-09-22 10:36:04 -07:00
Guoyu Wang
d957dbebea
Fix possible ios build break after update to Xcode 12 (#5246) 
* Fix possible ios build break after update to Xcode 12

* Address comments
 2020-09-22 07:42:54 -07:00
suffian khan
417929b049 jobs timeout .. 2020-09-21 21:51:59 -07:00
suffian khan
a6eb90472c try fix error on code coverage ci build 2020-09-21 21:51:59 -07:00
Sherlock
1478643215
Place Shape's output in CPU memory (#5245) 
Co-authored-by: Sherlock Huang <bahuang@OrtTrainingDev3.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net>
 2020-09-21 20:21:59 -07:00
Sherlock
038192bdb2
Place shape related compute nodes in CPU (#4940) 
* Place shape related nodes in CPU
* visit candidates by topological order
* Make CPU node placement a utility function
* skip placing on CPU if the data typs is float16 or bfloat16
 2020-09-21 17:10:39 -07:00
Changming Sun
0cb09374c6
Update BUILD.md for CUDA versions (#5239) 2020-09-21 15:28:53 -07:00
George Wu
3147bc00c3
update TensorRT docs (#5238) 
* doc updates TensorRT

* update

* update

* fix warning

* newline

* format
 2020-09-21 15:24:20 -07:00
Xueyun Zhu
55e4b5d302
add pipeline distributed training test (#5222) 
* add pipeline distributed training test

* fix max line length error in windows build

* function header indent

* fix

* fix flake8 error
 2020-09-21 14:35:01 -07:00
liqunfu
84c222126c
Deprecate testMNISTTrainingAndTestingOpset10 (#4927) 
Co-authored-by: liqun <liqun@OrtTrainingDev4.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net>
 2020-09-21 14:17:08 -07:00
Pranav Sharma
974b9bfc09
Allow sharing of initializers between sessions. (#5092) 
* Allow sharing of initializers between sessions.

* Allow sharing of initializers between sessions (2).

* Add test for C#

* Add test for C#; address PR comments

* Address PR comments
Moved AddInitializer logic to internal session options
Added tests for owned buffer
Clarified documentation
Fix bug where memory info and not device was getting compared

* Fix test

* Fix training build

* Add ver 5 end marker and ver 6 starter, add scenario and usage examples.
 2020-09-21 14:09:37 -07:00
Scott McKay
e0719a1073
Revert to using release SafeInt repo now that it supports a build with exceptions disabled. (#5233) 2020-09-22 06:29:28 +10:00
edgchen1
e9671e93f0
Fix TransposeScaleMatMul and MatMulScaleFusion issues (#5230) 
- Rename TransposeScaleMatMul back to TransposeMatMul for backwards compatibility
- Fix MatMulScaleFusion issues:
  - Add check for supported execution providers
  - Add check for supported MatMul input types
 2020-09-21 12:34:01 -07:00
Ye Wang
65740deb10
Fix a bug in EmbedLayerNorm fusion (#5150) 
* fix embedlayernorm bug

* review comments

* interim checkin

* review comments

* Fix core dump in MacOS

* remove unnecessary lines

* update document

* Update graph_utils.cc

* Update onnx_exporter.py

* resolve comments
 2020-09-21 12:26:14 -07:00
stevenlix
aefb2cc49b
Create profile for all dynamic shape input tensors (#5229) 2020-09-20 05:55:21 -07:00
Tiago Koji Castro Shibata
cd663d58f5
Fix WinML warnings (#5228) 2020-09-19 12:41:42 -07:00
Guoyu Wang
78a29aebbc
[ORT Mobile] ORT Minimal E2E CI (#5200) 
* Modify the ort minimal CI to ort minimal e2e ci
 2020-09-19 18:43:22 +10:00
Dmitri Smirnov
8ee4e8226e
Preserve relative order of the results and the tests. (#5225) 2020-09-19 00:45:44 -07:00
Weixing Zhang
b49f6a5e2c
using GPU_WARP_SIZE to make kernel portable between AMD and Nvidia GPU (#5173) 2020-09-18 14:56:16 -07:00
Suffian Khan
84589c7e05
Fuse softmax(a + b) in case of simple broadcast (#4937) 
* bias softmax kernel

* bias softmax kernel

* remove debug comments

* remove debug comment

* windows build doesnt handle unary minus on unsigned type

* int64 => int treated as error

* only support cuda

* add bias softmax fusion tests

* PR comments

* more PR comments

* use MLTypeCallDispatcher

* break function into pieces

* add loop unroll and add to list for inference as well

* use std::min and move operator==

* revert std::min (doesnt work ci pipeline) and fix int to size_t error

* pr comments

* fixes for windows ci

* fix for windows ci

* pr comments on consistency

* p_model_

* fix formatting and add anonymous namespace

Co-authored-by: suffian khan <sukha@OrtTrainingDev1.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net>
 2020-09-18 14:15:55 -07:00
Tang, Cheng
e0b49844e9
Provide option to let layernorm stash mean/var as fp32 or bfloat16 (#5215) 
* add option to set layernorm stash type

* bug fix

* fix merge error

* fix win build error
 2020-09-18 13:42:01 -07:00
Dmitri Smirnov
a90ab12589
Refactor onnx_test_runner (#5169) 
Refactor onnx_test_runner for better object ownership, code readability and maintainability.
 2020-09-18 13:19:35 -07:00
Ryan Hill
13318ab0d4
Remove invalid install line (#5219) 2020-09-18 11:58:40 -07:00
Shucai Xiao
a632dd2d3b
Amdmigraphx improvements (#5158) 
* code backup

* remove unnecessary log info

* code backup

* code backup

* merge changes from master branch

* code backup

* code backup

* merge changes from master branch

* code backup

* code backup for constant folding enhancement

* code backup

* include more scenarios for constant folding

* code backup

* remove unnecessary code

* remove unnecessary log information

* fix an error in comments

* update algorithm to do graph partition

* code backup

* remove unnecessary log information

* remove an unused function

* remove unnecessary changes
 2020-09-18 11:56:50 -07:00
Weixing Zhang
f91248e0cc
remove curand_generator_ related code since it is not used. (#5220) 2020-09-18 11:50:35 -07:00
KeDengMS
ce3b67e0cd
[Python] Move symbolic_shape_infer from nuphar to tools (#5162) 
* [Python] Move symbolic shape inference from nuphar to tools

* Fix PEP8 ERROR
 2020-09-18 09:31:06 -07:00
RRRachelllll555
f7c1e51810
Remove shape inference and fix save large model(>2g) issue (#5210) 
* remove shape inference and fix save large model problem

* remove unnecessary import

* refine code and add external format for quantize_qat

* remove initializers in tensors_to_calibrate

* small refine

Co-authored-by: t-yguo <t-yguo@microsoft.com>
 2020-09-18 08:46:31 -07:00
Scott McKay
c46a480306
Update conversion script and process to simplify creating ORT format models and a minimal build (#5217) 
* Update conversion script and process to simplify creating ORT format models and a minimal build.
 2020-09-18 18:49:54 +10:00
George Wu
1b61dfaf69
fix _WIN32 (#5218) 2020-09-18 00:23:17 -07:00
Pranav Prakash
f5df96256c
Fix order of returned values in quantize_weight_per_channel (#5205) 
Must match returned order of `quantize_inputs`
 2020-09-17 17:57:46 -07:00
liqunfu
f37e1292a1
--shm-size=1024m to fix nccl shared memory issue (#5214) 
* --shm-size=256m to fix nccl shared memory issue

Co-authored-by: liqun <liqun@OrtTrainingDev4.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net>
 2020-09-17 17:21:47 -07:00
Guoyu Wang
8156e0dd10
[ORT Mobile] Some updates to iOS/Android build settings (#5184) 
* Update android CI and build settings

* add build_java to arm64 also

* Add ios signing param

* fix a small build warning

* address pr comments
 2020-09-17 15:53:14 -07:00
Tracy Sharpe
8698157112
NCHWc optimizer fixes for quantized models (#5203) 
This updates the NCHWc transformer to not interfere with quantized convolution models, based on observations from internal models. The tensor type for MaxPool must be float. The input to GlobalAveragePool/GlobalMaxPool must be in NCHWc format.
 2020-09-17 09:52:21 -07:00
Pranav Sharma
d535894297
Add API to allow configuration of the global thread pools. (#5199) 2020-09-17 09:19:18 -07:00
Suffian Khan
e01e0b2e40
Fix softmax_warp_backward math when is_log_softmax = True and register LogSoftmax CUDA kernel (#5160) 
* register logsoftmax cuda kernel; fix logsoftmaxgrad cuda kernal; fix tests to invoke dispatch_softmax_*

* forgot to remove axis check

* add tests all axis

Co-authored-by: suffian khan <sukha@OrtTrainingDev1.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net>
 2020-09-17 07:15:25 -07:00
S. Manohar Karlapalem
584638e5d3
Corrects doc typos and formatting (#5201) 2020-09-17 01:25:19 -07:00
Zhang Lei
cd0386b649
MaxPool versioning in quantization tools. (#5194) 
MaxPool versioning in quantization tools.
 2020-09-16 22:52:24 -07:00
Ryan Hill
b11c106346
Remove almost all of the reinterpret_casts from the provider shared API (#5190) 2020-09-16 17:00:15 -07:00
Vincent Wang
c37472a1aa
Mixed Precision Transformer and Gradient Builder Refactor (#4892) 
* transform mixed precision before build gradient

* resolve comments

Co-authored-by: Vincent Wang <weicwang@microsoft.com>
 2020-09-17 02:44:50 +08:00
Tiago Koji Castro Shibata
f3f119a945
Use onecore umbrella lib in onecore builds (#5182) 
* delayload hack

* Skip tests

* Onecore uses onecore umbrella

* Uncomment tests

* cleanup

* Disable dev mode for WinML
 2020-09-16 10:46:27 -07:00
Tiago Koji Castro Shibata
1a2e289d2d
Fix nuget build (#5163) 
* Fix nuget content

* Revert "Fix nuget content"

This reverts commit e2cdcec4e39964c50eac2fb306c7a4bb84352443.

* Nuget packaging

* skip tests

* msbuild path

* Force msbuild version

* Workaround https://github.com/NuGet/Home/issues/7621

* cleanup
 2020-09-16 10:37:09 -07:00
Dmitri Smirnov
e6f85f338e
Refactor TensorAt, prepare for release (#5180) 
* Refactor TensorAt
  locations* must be const and int64_t since our dims are int64_t
  Remove unnecessary copy of locations.
  Remove unnecesary casting and C-casting. Simplify implementation.
  Add a check for string type.
  Make CXX api return T& to fully expose C API in C++, const std::vector& by value as it
  covers more ground and eliminate redundant copy.
  Eliminate inner loop, compute strides first.
 2020-09-16 10:20:45 -07:00
edgchen1
a20f8037f6
Install ssh in builder image, fix segfault in TrainingRunnerTest.Basic. (#5186) 2020-09-16 09:53:30 -07:00
Bowen Bao
400ac85565
Improve error message for FE model export checking (#5156) 2020-09-16 09:22:37 -07:00
Changming Sun
965e2b095d
Update MCR CUDA docker image to 10.2 (#5181) 2020-09-16 09:01:31 -07:00