onnxruntime/tools/ci_build/github/linux
PeixuanZuo 8ede2f139e
[ROCm] Optimize ROCm CI pipeline 2 (#16691)
- Set `KERNEL_EXPLORER_TEST_USE_CUPY=1` to replace numpy with cupy on
kernel explorer test.

KERNEL_EXPLORER_TEST_USE_CUPY=0 The CPU utilization is shown as below:

![image](https://github.com/microsoft/onnxruntime/assets/94887879/91724b78-0b4e-4cbd-ad88-83cad9976472)

KERNEL_EXPLORER_TEST_USE_CUPY=1 The CPU utilization is shown as below:

![image](https://github.com/microsoft/onnxruntime/assets/94887879/58239911-667c-4d5f-bb78-deca60d0266f)


- Use `Bash@3`.
- Update shell script.
2023-07-24 13:57:48 +08:00
..
docker [ROCm] Optimize ROCm CI pipeline 2 (#16691) 2023-07-24 13:57:48 +08:00
ort_minimal
tvm Fix TVM pipelines (#15653) 2023-04-26 09:55:05 +02:00
build_cuda_c_api_package.sh
build_linux_arm64_python_package.sh Fix python pipeline for AzureEP without using root (#16023) 2023-05-22 16:38:47 -07:00
build_yocto.sh
copy_strip_binary.sh Copy missing training header files to the package archive (#16119) 2023-05-30 16:45:40 -07:00
create_package.sh
extract_and_bundle_gpu_package.sh
java_copy_strip_binary.sh
java_linux_final_test.sh
run_build.sh Fix Linux MultiGPU TensorRT CI (#15697) 2023-04-26 10:01:33 -07:00
run_dockerbuild.sh Fix Linux MultiGPU TensorRT CI (#15697) 2023-04-26 10:01:33 -07:00
run_python_dockerbuild.sh Add compilation cache in 2 Linux CPU pipelines and refactor the Linux build step with cache (#15484) 2023-04-14 23:56:59 +08:00
run_python_tests.sh Unify cuda & trt version on few CIs (#15943) 2023-05-15 09:54:30 -07:00
test_custom_ops_pytorch_export.sh
upload_code_coverage_data.sh
upload_ortsrv_binaries.sh
yocto_build_toolchain.sh