* [NupharEP] Enable parallel schedule
* Update TVM with the fix to TVM threadpool to use OpenMP if possible
* Add parallel schedule when trying to vectorize
With this change, BERT squad perf on a 4-core (8 HT) CPU goes from 187ms to 150ms
* Address CR, docs and cmake update
* Doc fix
* Fix mkl
* Fix TVM windows build when using mklml
* update dockerfiles/README (#2336)
* Make elementwise op run 4 items per thread (#2335)
Description: Describe your changes.
Make elementwise op run 4 items per thread
unroll for loop to leverage ILP
remove unnessary N==0 check inside elementwise GPU kernel
Motivation and Context
Why is this change required? What problem does it solve?
It can improve the performance of GPU elementwise ops. ~2% performance gain on popular NLP bert model.
If it fixes an open issue, please link to the issue here.
* Add CUDA GatherElements kernel (#2310)
* Updates
* Update test
* Update
* Updates
* nits
* PR feedback
* Update
* Update
* PR feedback
* PR comments
* Update
* Fix build
* Fix build
* Nits
* Fix
* Layer Normalization Fusion (#2319)
basic layer normalization transform
* Add FastGelu Cuda Op for Gelu and Add bias fusion (#2293)
* Add FastGelu cuda op
* Add AddBiasGelu for experiment
* Revert "Add AddBiasGelu for experiment"
This reverts commit 5c1ee019858c657e6bb75887265cb85675626e5b.
* Add bias
* Add unit tests
* update comment
* update script
* fix build error
* update coding style
* update for CR feedback
Enable half2 optimization only when cuda arch >= 7.0
* move _Tanh to common.cuh
* implement CPU contrib OP Attention (#2333)
* Remove unused initializer from GraphProto as well as name_to_initial_tensor_ in CleanUnusedInitializers. (#2320)
* Remove unused initializer from GraphProto as well as name_to_initial_tensor_ in CleanupUnusedInitializers.
This means initializers that have been replaced during graph optimizations are not left in the GraphProto when we save an optimized model.
* Handle edge case where a model has an unused initializer with matching graph input by also removing the graph input.
* Use non-const iterators in std::find_if calls to make centos build happy.
* Nuget pipeline changes (#2305)
1. refactor the pipeline, remove some duplicated code
2. Move Windows_py_GPU_Wheels job to Win-GPU-CUDA10. We'll deprecated the "Win-GPU" pool
3. Delete cpu-nocontribops-esrp-pipeline.yml and cpu-nocontribops-pipeline.yml
4. In Linux nuget jobs, run "make install" before creating the package. So that extra RPAH info will be removed
* Cuda Reverse Sequence Op, maping types of same size using same template function. (#2281)
* Set ElementType to String type of node metadata, instead of byte[] (#2348)
* Set ElementType to String type of node metadata, instead of byte[]
* Fix spacing
* Introduce PrimitiveType into a Type System along with an integer constant (#2307)
Improve perf by avoiding GetType<T>() calls. Introduce MLTypeCallDispatcher to switch on Input Type. Add Tensor IsType<T>() fast method.
* Fix/test dim value of 0 handling in a couple of places (#2337)
* Update the CUDA Where implementation broadcasting logic to handle a dim with value of 0.
Add unit test
Also add unit test for unary op with dim value of 0
* Exclude ngraph from Where test with 0 dim.
* Openvino EP R3.1 onnxrt server (#2357)
* onnxrt server with OVEP
* onnxrt server with OVEP
* Update Dockerfile.server.openvino
* onnxrt server OVEP fix reviews
* onnxrt server OVEP fix reviews
* Implement cuda nonzero op. (#2056)
Implement cuda nonzero op.
* Direct use python numpy array's memory if already contiguous. (#2355)
* Direct use python numpy array's memory if already contiguous. This
could greatly improve performance for session with large input,
like big image 1920x1080 fastrcnn, 30~40% speed up could be achieved.
* Add test case enforce contiguous/non-contiguos numpy array as inputs.
* Add helper to create output to minimize binary size. (#2365)
Add ConstEigenTensorMap typedef so we don't unnecessarily const_cast the const input Tensor.
* fix builds enabling onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS (#2369)
* fix builds enabling onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS
* update
* Add Tracelogging for profiling (#1639)
Enabled only if onnxruntime_ENABLE_INSTRUMENT is ON
* test bidaf with nuphar for avx target (#2370)
increase nuphar test coverage a bit
* Fix a bug in TLS refcount that may destabilized CUDA CI (#2374)
* update output size calculation for resize (#2366)
* change how output size is calculated for resize op
* add tests for ver 10 resize
* Extend OneHot CPU kernel to support more types (#2311)
* Extend OneHot CPU kernel to support input int64_t, depth int32_t, output float
* Skip BERT before the test data fix is picked up
* Fix bug with Slice. Need to pass in flattened input dimensions so the initial offset into the input is calculated correctly. (#2372)
* Add opset 11 version of Split to CUDA ops (#2376)
Organize the CUDA ops definitions so all the opset 10 and 11 parts are together (same setup used for CPU ops)
* Layer Norm Fusion Fix (#2379)
* layer norm fusion fix
* Add input shape check in code and unit tests
* Fuse Add + Gelu (#2360)
Implement the transformer to fuse add + gelu
Implement the accurate kernel
* Skip layer norm transform (#2350)
* skip layer normalization transformer
* Another try to stabilize CUDA CI (#2383)
The root cause seems to be failure in CUDA dealloc when tear down. cudaFree return code was ignored before, so should the debug check.
* fix BUILD.md typo (#2375)
build.py: error: argument --config: invalid choice: 'RelWithDebugInfo' (choose from 'Debug', 'MinSizeRel', 'Release', 'RelWithDebInfo')
* Fixed compilation with ngraph (#2388)
* Fix reuse logic in allocation planner. (#2393)
* Fix reuse logic in allocation planner.
* PR comments
* Add helpful comments
* Don't allow reuse across string tensors.
* [NupharEP] Multiple optimizations (#2380)
Fuse transpose into MatMul
Implement Pow and constant scalar simplification
Vectorize ReduceMean
Improve symbolic shape inference
Minor updates for better debugging in fused function name
* Avoid using the default logger in the graph lib and optimizers (#2361)
1. Use the session logger if it is available.
2. Don't disable warning 4100 globally. We should fix the warnings instead of disabling it.
* Change CUDA implementation of Transpose to support all fixed size tensor types (#2387)
* Change CUDA implementation of Transpose to not use a typed kernel so we can support more types with minimum binary size.
Add support for 8, 16, 32 and 64 bit types.
Add unit tests.
Add method so the implementation can be called directly (will be used by CUDA Scan very soon).
* Disable TensorRT for MLFloat16 and int8 unit tests.
* Address PR comment and add support for calling cublas implementation if type is mlfloat16.
* Add opset 11 versions of the existing CUDA operators that had negative axis support explicitly added. (#2398)
* Add opset 11 versions of the existing CUDA operators that had negative axis support explicitly added.
* [NupharEP] force some low/zero cost ops to be inlined (#2409)
* fix cross compile bug (#2415)
* Minor optimization: if a node has already been placed, there's no need to find a kernel for it. (#2417)
* Add Reshape Fusion (#2395)
* Add reshape fusion
* Add some comments
* update comments
* update comment format
* update according to feedback
* update for recent logger change
* fix build error
* (1) Support both input and output edges in find path in graphutils
(2) Add a test case of only one constant initializer of Concat input.
(3) Refactor ReshapeFusion class to allow add more subgraph fusion in the future.
* fix error
* (1) loose constraint on initializer: non constant is allowed for reshape fusion.
(2) Change versions type to vector.
(3) Add logging.
(4) Return false when multiple output edges matched in FindPath. Add comments.
* only allow one direction (input or output) in FindPath
* [NupharEP] Update notebook and docker image (#2416)
Add BERT squad in Nuphar tutorial
Enhance speed comparsion readability
* Fix the issue in matmul_add_fusion (#2407)
Fix the issue in matmul_add_fusion
If Muatmul + Add has shape [K] * [K, N], reset it to [1, K] * [K, N] will make the output shape to [1, N] will also requires a reshape on the output.
Fix: just remove the shape reset to not fuse it.
Add a negative test case for matmul+add fusion
* feat(treeregressor): Update TreeEnsembleRegressor for type support (#2389)
Updates the `TreeEnsembleRegressor` to allow for `double`, `float`,
`int64`, and `int32` inputs to match the upstream specification.
Signed-off-by: Nick Groszewski <nicholas.groszewski@capitalone.com>
* onnxrt server documentation update (#2396)
* Added support for Pad-2 operator in OpenVINO-EP (#2405)
* Add CUDA If operator. (#2377)
* Add CUDA If operator.
Uses CPU operator for implementation.
By adding a CUDA version the inputs/outputs (with the exception of the 'cond' input) stay on GPU, and no other logic is required to avoid a copy to CPU across the control flow node.
* Improved documentation for onnxruntime::utils::SwapByteOrderCopy(), added precondition check.
* Fix the type constraints on CUDA If operator to exclude strings. (#2431)
* add Im2col<uint8_t> (#2438)
* Adjust codegen vectorization width from target (#2439)
* Adjust codegen vectorization width from target
* Add CUDA Scan operator. (#2403)
* Add Scan CUDA op.
Uses CPU implementation for logic.
Added some device specific functors for handling when data needs to be manipulated on a different device.
Added ability to override the materialization logic in the OrtValue slicer so DML can plugin their handling.
* Fix Windows GPU C API packaging pipeline failure (#2440)
Fix Windows GPU C API packaging pipeline failure (#2440)
* Correctly handle implicit inputs for fused nodes (#2390)
* Correctly handle implicit inputs for fused nodes
Previously, nuphar's partitioning function didn't include
node's implicit inputs into the inputs list of MetaDef, and hence
a crash was triggered in the onnx graph checker.
This commit fixed the issue. Furthermore, it also fixed a related
issue where we didn't add implicit inputs into
graph_inputs_excluding_initializers_ in Graph::SetGraphInputsOutputs.
the issue was that graph_inputs_including_initializers_ populated by
SetInputs (e.g. called by FunctionImpl::FunctionImpl) may contain
implicit inputs which were not of any node's initializers in the graph.
Because they were not part of any initializers, these implicit inputs
couldn't be visited by going through all nodes' inputs.
Consequently, they would *not* be added into graph_inputs_excluding_initializers_.
We fixed the issue by first copying the populated graph_inputs_including_initializers_
into graph_inputs_excluding_initalizers_, which then had both initializers and
non-initializers as its initial content. Later, we erase initializers from the
list. In this way, we can ensure all implicit inputs to remain in
graph_inputs_excluding_initializers_.
* refined comments and fixed duplicates
Address CR by revisiting comments in terms of implicit inputs
Also fixed an issue by skipping duplicates while copying inputs
from graph_inputs_including_initializers_.
* address CR
explain why we need to collect nodes' implicit inputs
* don't rely on pointer values for iterating std::set
Previously, openvino relied on iterating a set of NodeArg pointers
to construct inputs and outputs for a fused graph. It could cause
non-determinism. The reason was that although iterating std::set by
itself is stable, pointer values of NodeArgs may vary. Consequently,
we could end up visiting the set's elements in different orders for
different runs for the same test, which resulted in constructing
inputs (and outputs) with different orders to the fused graph.
For example, for the same test, we may have inputs [A, B] in some
runs but inputs[B, A] in others.
Let's use std::string as the key type to avoid such nondeterminism.
This commit also added implicit inputs into meta->inputs while returning
the capability from the openvino provider.
* Fixed another latent issue in openvino's GetCapability function
The issue was that we couldn't simply erase fused_inputs and fused_outputs
while iterating the nodes. For example, an output NodeArg may have multiple
uses, and it's wrong if we erase it from fused_outputs when we encounter only
one of its uses as input.
* Remove DeviceAllocatorRegistry class (#2451)
Remove DeviceAllocatorRegistry class
* CSharp api and test for loading custom op shared library (#2420)
- Added C-API test for loading custom op shared lib.
- Made some changes in C++ api header and C-api implementation to get it working.
- Added C# API and corresponding test for loading custom op shared library.
* Parallel Gelu with ParallelFor (#2399)
Parallel Gelu to get better performance for Gelu
* Clean up build.py (#2446)
* Pull the latest image before running docker build
* Fuse SkipLayerNorm with Bias (#2453)
Fuse SkipLayerNorm with Bias
* Allow more than one invocation of CreateEnv in the same process. (#2467)
* Allow more than one invocation of CreateEnv in the same process.
* Fix centos build
* Symbolic shape inference improvements: (#2460)
* Symbolic shape inference improvements:
- add a mode to guess unknown ops' output rank
- add support for GatherND
- add support for If
- fix a bug in get_int_values when then tensor rank > 1D, by treating it as no sympy data
- add symbol to literal merge when ONNX silently merges dims
- fix a bug in Concat when input dim is 0
- fix a bug in ConstantOfShape that computed dim is not updated
- add support for dynamic shape in ConstantOfShape
- fix a bug in Loop output shape that loop iterator dim is not inserted at dim 0
- add support for dynamic padding in Pad
- add support for dynamic shape in Reshape
- add support for Resize with opset > 10, by treating output dims as dynamic
- fix a bug in Slice when starts/ends are dynamic
- restrict input model to opset 7 and above
- make output model optional to avoid disk write when testing
Run model tests for symbolic shape inference
Reduce 2GB docker image size of nuphar
* add additional test data set for nuget pipeline (#2448)
* add SAS token to download internal test data for nuget pipeline
* update azure endpoint
* fix keyvault download step
* fix variable declaration for secret group
* fix indentation
* fix yaml syntax for variables
* fix setting secrets for script
* fix env synctax
* Fix macos pipeline
* attempt to add secrets to windows download data
* fix mac and win data download
* fix windows data download
* update test data set url and location
* Pass flags to build/test WinML in CI
* Add initial CMake config for unit tests in WinML
* Set winml_unittests standard to C++17
* Add WinML API tests and port them to googletest
* Install WinML test collateral
* Add LearningModelSessionAPITests ported to googletest
* Fix WinML test files encoding
* Add GPU tests
* Add parameterized test, skip GPU tests
* Enable precompiled header
* Remove unused code and collateral
* Remove brand images
* Add dllload.cpp
* Remove images not used in API tests
* Add LICENSE.md to image collaterals
* Add models with licenses
* Remove FNS Candy tests
* Add API test models
* Add ModelInSubdirectory
* Install collaterals post-build with copy_if_different, split common lib
* fix warnings
* Link to gtest_main
- Added C-API test for loading custom op shared lib.
- Made some changes in C++ api header and C-api implementation to get it working.
- Added C# API and corresponding test for loading custom op shared library.
* add option to enable winml telemetry
* add option to enable winml telemetry
* clean logs while developping
* clean the log of GUID
* compile onnxruntime_common with winml telemetry
* use option for use_telemetry
* rename option winml_use_telemetry to onnxruntime_use_telemetry
* little change
* onnxrt server with OVEP
* onnxrt server with OVEP
* Update Dockerfile.server.openvino
* onnxrt server OVEP fix reviews
* onnxrt server OVEP fix reviews
1. refactor the pipeline, remove some duplicated code
2. Move Windows_py_GPU_Wheels job to Win-GPU-CUDA10. We'll deprecated the "Win-GPU" pool
3. Delete cpu-nocontribops-esrp-pipeline.yml and cpu-nocontribops-pipeline.yml
4. In Linux nuget jobs, run "make install" before creating the package. So that extra RPAH info will be removed
add onecoreuap_apiset.lib in order to avoid linking against kernel32.lib etc and violating our OS layering requirements.
We linked against onecoreuap_apiset.lib in VB so we will continue doing this, but I am still unsure why not to link against onecore instead since that is where we ship. However, since Sheil is the owner of this code we will wait to discuss with him before changing anything.
* Guard unused parameter
Guard unused parameter for Linux Arm and other cases.
* Add ACL (Arm Compute Library) execution provider
Add a new execution provider targeting Arm architecture based on Arm Compute Library.
Validated on NXP i.MX8QM CPU with ResNet50, MobileNetv2 and VGG models.
All unit tests are passing.
Comparative performance improvements for ResNet50v1 model obtained with
onnxruntime_perf_test:
A72 2xA72 A53 4xA53
ACL vs CPU 16% 9% 21% 13%
Usage documentation available in ACL-ExecutionProvider.
* Fix eigen unused parameter
Fix eigen unused parameter error for Arm cross-compilation.
* Refine optimizers
* Address PR comments
* Changes from PR comments and discussion.
* Fixed signed/unsigned mismatch
* Address PR comments
* Address PR comments
* Fix linux build
* Fix issue with mkldnn logic.
* Turn off optimizers by default for operator unit tests.
* Handle edge case of graph with no nodes in partitioner so all execution providers don't need to.
* Comment out change to turn off optimizers for unit tests. Add details on what needs to be done to re-enable.
This change adds a new execution provider powered by [DirectML](https://aka.ms/DirectML).
DirectML is a high-performance, hardware-accelerated DirectX 12 library for machine learning on Windows. DirectML provides GPU acceleration for common machine learning tasks across a broad range of supported hardware and drivers.
The DirectML execution provider is capable of greatly improving evaluation time of models using commodity GPU hardware, without sacrificing broad hardware support or requiring vendor-specific extensions to be installed.
**Note** that the DML EP code was moved verbatim from the existing WindowsAI project, which is why it doesn't yet conform to the onnxruntime coding style. This is something that can be fixed later; we would like to keep formatting/whitespace changes to a minimum for the time being to make it easier to port fixes from WindowsAI to ORT during this transition.
Summary of changes:
* Initial commit of DML EP files under onnxruntime/core/providers/dml
* Add cmake entries for building the DML EP and for pulling down the DirectML redist using nuget
* Add a submodule dependency on the Windows Implementation Library (WIL)
* Add docs under docs/execution_providers/DirectML-ExecutionProvider.md
* Add support for DML EP to provider tests and perf tests
* Add support for DML EP to fns_candy_style_transfer sample
* Add entries to the C ABI for instantiating the DML EP
In ORT, there is only 3 cuda stream: default, HtoD, DtoH. And both HtoD and DtoH are non-blocking stream. Thus, per-thread stream mode doesn't have any benefit.
I also tried in multiple thread env and the legacy mode is also better than per-thread model.
Below is the perf of a 3 layer bert on v100. Unit is ms:
batch size 1:
concurrency | c=1 | c=2 | c=4
legacy | 0.54 | 1.17 | 2.68
per-thread | 0.66 | 1.37 | 2.86
batch size 4:
concurrency | c=1 | c=2 | c=4
legacy | 1.1 | 2.22 | 4.6
per-thread | 1.21 | 2.44 | 4.98
batch size 64:
concurrency | c=1 | c=2 | c=4
legacy | 8.09 | 16.13 | 32.37
per-thread | 8.18 | 16.26 | 32.45
* Adjust ngraph cmake files to onnx 1.5.0
* Enable LSTM reverse direction mode in nGraph EP
* Enable full support for the Split op in nGraph EP
* Revert "Disable the unsigned input Shrink op tests for nGraph until the next update"
This reverts commit 257b42a55bdd98f804d4846868542b8e3aeb4b4e.
* Enable Gather and remove unused subgraph attribute
* Remove the unused param from AppendClusterToSubGraph
* Fix for the incorrect onnx opset version
* Use the r0.26 release branch before the tag is created
* Enable the quantizelinear and dequantizelinear for NGEP
* Use the v0.26.0-rc.2 tag in ngraph.cmake
* Add skip for modes others than default in Pad operator
* Reenable negative axis tests for ngraph
* Use temporary ngraph version
* Use branch name instead of SHA for temporary ngraph branch
* Use ngraph v0.26.0-rc.4
* Remove patch for missing symbol in MKLDNN
* Use MKLDNN 1.0 in ngraph
* Exclude the Pad op for opsets greater than 10
* Disable quantizelinear and dequantizelinear tests for ONNX 1.5.0
* Fix the onnx-headers related compilation errors
* ONNX libs linking fix
* Use a tag for ngraph and support more Pad modes
* Use the v0.26.0 release tag for nGraph
* Update ngraph to RC8 - bigobj flag for Windows builds
* Fix the MKLDNN constexpr error on Windows
* save status: add tiling layout; add avx512 skylake cpuid info
* unit tests and matmul integer model passed on skylake, need to verify model
* save commit before update master
* fix check
* address comments
