saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-05-18 21:21:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
234 commits 2856 branches 57 tags 1.3 GiB
Commit graph

234 commits

This branch
This branch
All branches
Author SHA1 Message Date
Changming Sun
e97caa7787 change mkldnn so path (#210) 2018-12-19 14:45:57 -08:00
ashku-ms
84231ba003
support hyperbolic ops (#223) 
* support hyperbolic fns

This commit adds support for sinh and cosh. Support for hyperbolic inverses is not available in Eigen yet.

* Make constructors explicit

* remove tests from exclude list

* Revert "remove tests from exclude list"

This reverts commit 2112a30b57d5a899991de4847e948e700a44e85d.

* remove test names from excluded list

* remove tanh since its already implemented
 2018-12-19 14:23:09 -08:00
Edward Chen
e63572c1f3 Updated ArrayFeatureExtractor op to retain old output shape behavior. 2018-12-19 14:12:55 -08:00
Tang, Cheng
94f8f2b05c placeholder for internal contrib ops (#219) 2018-12-19 13:54:12 -08:00
KeDengMS
b9cc134576
Make sure tensor sizes are 64-byte aligned (#222) 
This helps reduce misaligned access violation
 2018-12-19 13:45:04 -08:00
Bowen Bao
4d010fb1ea Add null check before calling node.op_->Deprecated(). (#211) 2018-12-19 11:43:36 -08:00
Changming Sun
ac3a081ec5 Enable release build in Windows CI pipelines (#220) 2018-12-19 11:12:43 -08:00
Jesse Benson
0248390e4d Add support for checking for F16C support (https://en.wikipedia.org/wiki/F16C). (#212) 2018-12-19 09:25:25 -08:00
Scott McKay
ab350fa4c7
Re-structure the inference session initialization to (#217) 
- apply any transforms to the main graph and any subgraphs first
  - call Graph::Resolve() once on the main graph, which will recurse into the subgraphs
    - previously it was called after the transform on each subgraph, which results in it traversing up to the main graph to call resolve, and that resolve call recursing into all subgraphs every time.

This avoids lots of unnecessary Graph::Resolve calls, and prevents subgraphs from being broken by SessionStateInitializer::InitializeAndSave calling graph_.CleanAllInitializedTensors() prior to final Graph::Resolve call. If a subgraph has optional inputs the backing initializers were removed by CleanAllInitializedTensors causing the next Resolve to incorrectly turn them into required inputs.
 2018-12-19 18:56:35 +10:00
Scott McKay
334e329642
Increment/decrement UseCount for outputs so that we don't prematurely free a re-used output that is used for a dead output (output with zero users). (#214) 2018-12-19 15:34:37 +10:00
Pranav Sharma
39f47f86ee
Adding the include folder for the C Windows pkg. (#198) 
* Adding the include folder for the C Windows pkg.

* Add import lib to the pkg

* Disable csharp pretrained tests temporarily
 2018-12-18 21:04:42 -08:00
Scott McKay
beb326f00e
Simplify logic around creating relationship between nodes for implicit NodeArg usage. Allows using an initializer from multiple levels up to not fail. We would need to accumulate a list of initializers from all levels up otherwise, and doing so doesn't add any value. (#200) 
Improve a comment to clarify when the parent graph NodeArg lookup kicks in.
 2018-12-19 09:25:42 +10:00
Changming Sun
dc8b37f4c4
update onnx (#209) 
* update onnx
 2018-12-18 14:50:28 -08:00
Ke Zhang
37b74c771a
add gemmlowp as submodule. (#206) 2018-12-18 13:57:53 -08:00
Changming Sun
c0ec7d56c4
Clean up garbage files (#208) 2018-12-18 13:51:01 -08:00
jignparm
d4131a31d9
Disable csharp pretrained tests temporarily (#207) 2018-12-18 21:35:23 +00:00
edgchen1
0aa1b54aaa
Updated build.py to support relative CMake/CTest paths and did some minor cleanup. (#205) 2018-12-18 13:23:32 -08:00
Ryan Hill
773114a4f1
More C header naming changes (#202) 
* More Ort prefix changes for consistency

* Fix C# methods

* More C# fixes
 2018-12-18 11:39:46 -08:00
jignparm
0287019c1d C# Gpu : Minor updates to exception message (#201) 
* Minor updates to exception message

* update models folder to new location

* update copy to preservenewest
 2018-12-18 00:30:27 -08:00
edgchen1
71c56b6d7c Fix array feature extractor out of bounds access issue (#194) 
* Fixed out of bounds access in ArrayFeatureExtractor.

* some cleanup

* Updated tensor_shape.h comments.

* Updated macro name.

* Added copy assignment, move assignment/ctor to TensorShape.

* Removed i64 literal suffix.

* Fixed test.

* Fixed type of x_num_dims.
 2018-12-18 00:30:07 -08:00
Changming Sun
8c5d105557 fix typo 2018-12-17 18:33:36 -08:00
KeDengMS
b0f27ba0a7
Allow using MKLML header/libs when use_mklml is specified (#178) 
Allow using MKLML header/libs when use_mklml is specified
 2018-12-17 14:41:42 -08:00
Randy
d0544a8082
Rashuai/gathernd op (#170) 
* define gather_nd op

* add test cases

* add test file

* refactor the code and doc

* add test cases

* fix win compile err

* fix win compile err

* adjust indent

* make constructor explicit

* add coment

* remove templates

* remove wrong def

* migrate macros

* fix an issue in shape inference
 2018-12-17 13:47:20 -08:00
Ke Zhang
82d04412a0
Kezhan/partition logic update (#164) 
* add check before fusing sub-graph in greedy partitioning

* update the partitioning logic to 1) not fuse sub-graph if inner nodes were assigned 2) avoid resolving graph after each provider capability checking and assignment.

* resolve conflicts
 2018-12-17 13:30:29 -08:00
Ke Zhang
5a8acd7da8
directly updating dst arg with src arg when adding an edge. no need updating the type and shape info. (#195) 2018-12-17 13:30:16 -08:00
jignparm
5fd9024139
[WIP] Initial checking for CSharp GPU support (#176) 
* Initial checking for CSharp GPU support

* Enabled C# for GPU build

* Update Onnxruntime to Ort

* Add runtime check for cuda dlls for windows

* Update pretrained model test, for models where name!=model.onnx

* lowered tolerance for float checks to pass new models

* ignore extra ._resnet34v2.onnx file in pretrained test
 2018-12-17 21:18:48 +00:00
Changming Sun
b418adff42 Disable test data downloading by default (#193) 2018-12-17 11:56:29 -08:00
xkszltl
383315a7e0 Upgrade Eigen to 3.3.7 (#185) 2018-12-16 00:26:08 -08:00
Tongliang Liao
16b3a9fd63 onnxruntime_CUDA_HOME should only be used when available. 2018-12-16 00:25:12 -08:00
Du Li
6cd935e184
Duli/fusedconv (#183) 
* Adding missing attribute *alpha* to FusedConv schema.
 2018-12-14 16:00:18 -08:00
Du Li
7560538c94
Revert "Adding missing attribute *alpha* to FusedConv schema." (#182) 
This reverts commit 3f4e62a49c.
 2018-12-14 15:39:34 -08:00
Du Li
3f4e62a49c Adding missing attribute *alpha* to FusedConv schema. 2018-12-14 15:17:16 -08:00
Changming Sun
55a0527640 Fix a warning in gcc 7 2018-12-14 15:16:52 -08:00
Ryan Hill
11b369a864
Abbreviate ONNXRuntime as Ort in all of our public APIs (#175) 
Applies to all public headers and macros, plus many internal ones. There are still some internal things with OnnxRuntime in the name, but this fixes all public functions & macros.
 2018-12-14 14:54:23 -08:00
Tracy Sharpe
47551da994
Optimize Tanh/Sigmoid activations (#162) 
* optimized tanh/sigmoid

* fix /W4 warnings from alternate build environment

* use MLAS for tanh/sigmoid

* fix my broken C++ templates

* add x86_64 files
 2018-12-13 22:53:40 -08:00
Changming Sun
afd831ae9e Remove unnecessary const 2018-12-13 21:51:28 -08:00
Hector Li
d6abc38182 MurmurHash3 operator (#174) 
* MurmurHash3 operator
 2018-12-13 19:52:00 -08:00
Changming Sun
fb021b9002 delete azure-pipelines.yml 2018-12-13 18:38:59 -08:00
Weixian
2ffaa8a185 optimize unordered_map (#166) 2018-12-13 16:02:50 -08:00
Changming Sun
0aaaf4663d Update data download script (#171) 
* Add cache dir

* update

* disable csharp

* update

* Revert "disable csharp"

This reverts commit e1a80f3272f7e7881f3081a91467756d2769fdf4.

* add output

* update
 2018-12-13 14:46:59 -08:00
Ryan Hill
d419170c7b
Rename ONNXRUNTIME_API_STATUSCALL to ONNXRUNTIME_API_CALL since it's … (#165) 
* Rename ONNXRUNTIME_API_STATUSCALL to ONNXRUNTIME_API_CALL since it's just for calling convention, there's no status return value implied.

* Missed a function
 2018-12-13 14:42:51 -08:00
edgchen1
c5a0119d42 Added Environment::IsInitialized() and added check to InferenceSession constructor. (#169) 2018-12-13 13:34:49 -08:00
Ryan Hill
34175826df
Ryanunderhill/pad fix (#151) 
* Fix pad bug when there is negative padding

* Add in dimension count check
 2018-12-13 10:44:01 -08:00
Ke Zhang
1d32aa98f8
add int16/uint16 in quantization for cntk quantized speech model. (#163) 2018-12-12 18:08:13 -08:00
Ryan Hill
30ab01e7be
Ryanunderhill/concat neg axis (#152) 
* Add negative axis support to concat

* Unit test bug
 2018-12-12 18:01:28 -08:00
Changming Sun
cfd3b4c606
remove extra const (#159) 2018-12-12 16:53:45 -08:00
Changming Sun
ed98d3d653
Add some nullptr checks (#160) 2018-12-12 16:53:35 -08:00
Changming Sun
618cc51754
Update onnx_backend_test_series.py (#146) 
* Update onnx_backend_test_series.py

* Update BUILD.md
 2018-12-12 16:25:16 -08:00
Changming Sun
ec71f363f7
Fix a bug in parallel executor and enable test for it in CI (#157) 
* Fix a bug in parallel executor and enable test for it in CI
 2018-12-12 15:52:36 -08:00
Changming Sun
1ed72251d0
A tiny change to onnxruntime::Model (#158) 
* Remove unnecessary const

* update
 2018-12-12 15:33:20 -08:00