Add SpaceToDepth and DepthToSpace CUDA NHWC Ops #19646
Conversation
mtavenrath
force-pushed
the
nhwc_depth_to_space
branch
3 times, most recently
from
b683eaf to
521fbbb
Compare
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-python-checks-ci-pipeline,onnxruntime-binary-size-checks-ci-pipeline,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 9 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
mtavenrath
changed the title
mtavenrath
changed the title
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
1 similar comment
|
Azure Pipelines successfully started running 10 pipeline(s). |
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
1 similar comment
|
Azure Pipelines successfully started running 10 pipeline(s). |
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
1 similar comment
|
Azure Pipelines successfully started running 10 pipeline(s). |
… cuda_nhwc_kernels
mtavenrath
force-pushed
the
nhwc_depth_to_space
branch
from
8330ce3 to
ca45425
Compare
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
1 similar comment
|
Azure Pipelines successfully started running 10 pipeline(s). |
|
/azp run Windows ARM64 QNN CI Pipeline,Windows x64 QNN CI Pipeline,Windows CPU CI Pipeline,Windows GPU CI Pipeline,Windows GPU TensorRT CI Pipeline,ONNX Runtime Web CI Pipeline,Linux CPU CI Pipeline,Linux CPU Minimal Build E2E CI Pipeline,Linux GPU CI Pipeline,Linux GPU TensorRT CI Pipeline |
|
/azp run Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models,Android CI Pipeline |
|
/azp run iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 2 pipeline(s). |
|
Azure Pipelines successfully started running 10 pipeline(s). |
|
/azp run ONNX Runtime Web CI Pipeline,Linux OpenVINO CI Pipeline,Linux QNN CI Pipeline,MacOS CI Pipeline,orttraining-amd-gpu-ci-pipeline,orttraining-linux-ci-pipeline,orttraining-linux-gpu-ci-pipeline,orttraining-ortmodule-distributed,onnxruntime-binary-size-checks-ci-pipeline,Big Models |
|
/azp run Android CI Pipeline,iOS CI Pipeline,ONNX Runtime React Native CI Pipeline |
|
Azure Pipelines successfully started running 10 pipeline(s). |
### Description - Adding CUDA NHWC support for SpaceToDepth and DepthToSpace - Add a new test which verifies that swizzling SpaceToDepth swizzling for the H axis is correct. - If CUDA NHWC is enabled, run all tests on the CUDA EP with NHWC as well. ### Motivation and Context Adding more NHWC operations to avoid layout transformations when using the CUDA EP for more efficiency.
| @@ -692,7 +692,7 @@ TEST(UpsampleOpTest, NhwcUpsampleOp4D1CBilinearTest) { | |||
| // TensorRT: results mismatch | |||
| // ROCm: results mismatch | |||
| test.Run(OpTester::ExpectResult::kExpectSuccess, "", | |||
| {kCudaExecutionProvider, kTensorrtExecutionProvider, kRocmExecutionProvider}); | |||
| {kCudaExecutionProvider, kCudaNHWCExecutionProvider, kTensorrtExecutionProvider, kRocmExecutionProvider}); | |||
There was a problem hiding this comment.
If we are going through this approach for testing - should we get rid of the tests (or bring them over here) in tests/providers/cuda/nhwc ?
There was a problem hiding this comment.
The CPU tests tests against all backends except for the excluded ones - and those have been excluded for a reason. We should do the reverse, move cuda/nhwc tests over to the cpu tests for test cases which are not yet covered and remove the specialized nhwc tests.
There was a problem hiding this comment.
I agree. We should "unify" (i..e) pick up tests from the nhwc tests over to the cpu tests (after de-deuplication) and get rid of the nhwc tests.
Description
Motivation and Context
Adding more NHWC operations to avoid layout transformations when using the CUDA EP for more efficiency.