-
Notifications
You must be signed in to change notification settings - Fork 2.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add amd migraphx execution provider to onnx runtime #2929
Conversation
Thanks for the contribution! |
Would you be able to test against the models contained in the zip file? Those are models we test against in our CI pipelines using onnx_test_runner. |
why are updates showing up in the tvm and onnx-tensorrt modules? |
This a mismatch of the commit used by these submodules between our branch and the master branch. Fixed that. |
@@ -743,6 +744,14 @@ if (onnxruntime_USE_TENSORRT) | |||
endif() | |||
endif() | |||
|
|||
if (onnxruntime_USE_MIGRAPHX) | |||
if (WIN32) | |||
message(FATAL_ERROR "MIGraphX does not support build in Windows!") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
are there plans to eventually support in Windows?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We currently dont have any plans.
After trying all models, MIGraphX EP can run all cases by either falling back to CPU or run correctly on MIGraphX. |
/azp run Linux CPU CI Pipeline, Linux CPU x64 NoContribops CI Pipeline, Linux GPU CI Pipeline, Linux GPU TensorRT CI Pipeline, MacOS CI Pipeline, MacOS NoContribops CI Pipeline, Win CPU x64 NoContribops CI Pipeline, Win CPU x86 CI Pipeline, Windows CPU CI Pipeline |
Azure Pipelines successfully started running 9 pipeline(s). |
/azp run Windows GPU CI Pipeline, Windows GPU TensorRT CI Pipeline |
Azure Pipelines successfully started running 2 pipeline(s). |
/azp run MacOS CI Pipeline |
Azure Pipelines successfully started running 1 pipeline(s). |
Azure Pipelines successfully started running 5 pipeline(s). |
windows build is failing because of failing python pep8 check (which is new) |
Thanks. Just fixed it, hope it can pass this time. |
/azp run Linux CPU CI Pipeline, Linux CPU x64 NoContribops CI Pipeline, Linux GPU CI Pipeline, Linux GPU TensorRT CI Pipeline, MacOS CI Pipeline, MacOS NoContribops CI Pipeline, Win CPU x64 NoContribops CI Pipeline, Win CPU x86 CI Pipeline, Windows CPU CI Pipeline |
/azp run Windows GPU CI Pipeline, Windows GPU TensorRT CI Pipeline, orttraining-linux-ci-pipeline, orttraining-linux-gpu-ci-pipeline, orttraining-mac-ci-pipeline |
Azure Pipelines successfully started running 7 pipeline(s). |
Azure Pipelines successfully started running 5 pipeline(s). |
/azp run Windows CPU CI Pipeline |
Azure Pipelines successfully started running 1 pipeline(s). |
/azp run Linux CPU CI Pipeline, Linux CPU x64 NoContribops CI Pipeline, Linux GPU CI Pipeline, Linux GPU TensorRT CI Pipeline, MacOS CI Pipeline, MacOS NoContribops CI Pipeline, Win CPU x64 NoContribops CI Pipeline, Win CPU x86 CI Pipeline |
/azp run Windows GPU CI Pipeline, Windows GPU TensorRT CI Pipeline, orttraining-linux-ci-pipeline, orttraining-linux-gpu-ci-pipeline, orttraining-mac-ci-pipeline |
Azure Pipelines successfully started running 6 pipeline(s). |
Azure Pipelines successfully started running 5 pipeline(s). |
@snnn is the static analysis failure a known issue? i don't think it's due to changes in this PR |
Yes. Please update the branch with the latest master |
@scxiao please merge latest master to fix the windows cpu CI build failure (due to static analysis failure) |
Thanks for the info. I just merged that. Please try again. |
/azp run Linux CPU CI Pipeline, Linux CPU x64 NoContribops CI Pipeline, Linux GPU CI Pipeline, Linux GPU TensorRT CI Pipeline, MacOS CI Pipeline, MacOS NoContribops CI Pipeline, Win CPU x64 NoContribops CI Pipeline, Win CPU x86 CI Pipeline |
/azp run Windows GPU CI Pipeline, Windows GPU TensorRT CI Pipeline, orttraining-linux-ci-pipeline, orttraining-linux-gpu-ci-pipeline, orttraining-mac-ci-pipeline |
Azure Pipelines successfully started running 6 pipeline(s). |
Azure Pipelines successfully started running 5 pipeline(s). |
/azp run Windows CPU CI Pipeline |
Azure Pipelines successfully started running 1 pipeline(s). |
Description:
This adds an initial execution provider using MIGraphX which supports the AMD ROCm stack.
As a first step, we have hardcoded the operators supported. In the future, we plan to update this to use an API from MIGraphX directly.
This also uses the C/C++ API from MIGraphX, as this will be more stable. As this hasn't been released as part of the ROCm distribution yet we build this version in the dockerfile. In the future we can switch to just pulling down the debian package from ROCm.
Motivation and Context
To support AMD GPUs in onnx runtime and take advantage of MIGraphX's graph optimizations.