* Arm64 Depthwise Convolution 3x3.
* Add 5x5 intrinsic dwqconv for arm64
* rebase to master, remove no-need logic after arm64 convsym enabled.
* Some more adjustment on the instrunction pipeling.
* Add specific test cases.
* Fix test dimension too small.
* Fix build warning as error on some CI.
* better format, etc.
* add use_tensorrt build option
* Add use_tensorrt to running tests
* add use_tensorrt for Windows
* make trt ep to skip backend test
* make trt ep to skip backend test
* Fix bug
* Add/Modify description
* modify for debug
* swtich pool to test
* modify to debug
* modify to debug
* add vobersity
* refine the code
* refine the code
* refine the code
* fix flake8 warning
* refine the code
* add pre_load check for trt as well as add cupti lib to cuda depedencies
* modify script to make trt build path the same as cuda
* show error message when user wants to run TensorRT but TensorRT is not installed in the env
* fix bug
* fix bug
* add trt lib for manylinux
* include cuda_dependencies for trt
* rewrite the condition to throw exception
* make code more compact
* Add intermediate header between the ORT code and pybind11 to workaround an issue with VS2022 debug builds by making sure corecrt.h is included first.
This avoids the _STL_ASSERT macro being defined in an incompatible way for a debug build by pybind including the python headers with _DEBUG temporarily undefined .
See #9735 for details.
* register custom symbolic for einsum
* bugfix for case needs permute at the end
* refactor
* refactor equation parser
* support new case, use ReduceProd
* optimize perf and graph
* remove some Gather node
* add more ut, fix gemm trans fusion
* Update required operators for prebuilt package to add opsets 14 and 15.
Add helper script to check if the prebuilt package will support the model and if not why not.
* Add support for multiple opsets being specified on a single line in the required operators config. This makes it easier to update the pre-built package config.
It's also required for validation tools to work as they only have a single opset from the model and not per-operator opsets. If we only list the incremental ops we could merge in the ops from the previous opset, but that wouldn't give a way to drop an operator from being supported.
Left the info on which ops changed though so we have a better feel for the cost of supporting each opset.
* Added checks for Hetero/Multi
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Remote Context Plugin
* changes for IO Buffer plugin
* erronous couts added
* erronous entry rectified
* Set the Openvino OP Buffer also as output
* Enable AUTO plugin in OpenVINO EP
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Remote Context Plugin
* changes for IO Buffer plugin
* erronous couts added
* erronous entry rectified
* Added checks for Hetero/Multi
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Set the Openvino OP Buffer also as output
* Enable AUTO plugin in OpenVINO EP
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Please commit error message and rectification of param.context
* Alignment fixed
Signed-off-by: MaajidKhan <n.maajidkhan@gmail.com>
* Changed the string to OpenVINO_GPU
* hanged OpenVINO to to OpenVINO_CPU
* Onnxruntime updated API for memory location
* Removing Duplicate LOG Error
* Tensor.h removed DeviceType function. Updated comment
* API Comments updated
* Removing changes to Provider Indo
* Erronous commit
* Removing Extra logs
* Merge CMAKE
* Not copy from a local location
* Duplicate Entry
* Remove extra line
Co-authored-by: MaajidKhan <n.maajidkhan@gmail.com>
Adding ARM64 depthwise convolution kernel for symmetric quantization
Motivation and Context
Two improvements against current kernel code :
1. Signed int8 based instructions, no need to extend from 8b to 16b before multiplication.
2. Unrolled loop with manual software pipelining
Co-authored-by: Chen Fu <fuchen@microsoft.com>
* Only serialize runtime optimization records container if non-empty.
* Remove runtime optimizations from onnxruntime/core/flatbuffers/schema/README.md as it's not completely implemented yet.
* Disable partial runtime optimization implementation by default.
Fix Resize float16 ROI tensor. The ROI tensor is (and always has been) handled by the CPU, but we didn't register that type, nor have a switch case for it.
"NVIDIA shared a model with Resize that falling back to CPU due to Float16 resize with Float16 ROI."
```
te.exe OnnxConformanceTests.dll /unicodeOutput:false /name:*Resize* /p:"match=Resize*"
Summary: Total=79, Passed=76, Failed=0, Blocked=0, Not Run=0, Skipped=3
```
Related work items: #33939079
* schema change
* cc channges
* remove temp debug code
* Adding fbs namespace to session_state_flatbuffers_utils.h
* Add fbs namepsace to all ort format utils
When the pattern Sum(Gemm(A, B), C) exists, we can convert it to
Gemm(A, B, C), assuming that C the output of the original Gemm is
not used elsewhere, and this change does not break broadcasting.
ORT format model runtime optimization implementation is in progress.
This change adds a build.py option to disable the partial runtime optimization implementation, adds CI builds to test it, and disables runtime optimizations in mobile package builds.
* Construct valid graphs for ONNX checker for IR version < 4.
Previously the constructed graph was not guaranteed to have its
initializers be a subset of its inputs, which is required for IR
version < 4. This resulted in spurious failures.
Fixes#9663
implement dynamicquantizelinear in DNNL EP
add debug log in EP for operator coverage
block gpu elementwise op with 5 dims or more
Signed-off-by: Wang <zhaoyang.wang@intel.com>
* Fix#9671 by running the level 1 rewrite rules first and allowing the transpose optimizer to run multiple times to ensure it completes in level 1.
Removed unnecessary call to GenerateRuleBasedGraphTransformer as there are no level 2 rewrite rules.
