huggingface / huggingface/optimum-amd

[MI300X] warning: failed to meet occupancy target given by 'amdgpu-waves-per-eu' in

Open
#163 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Jupyter Notebook
Stars
101
Forks
21
Avg merge
58m
Merged PRs (30d)
1

Description

When I am compilling onnxruntime-rocm following the steps in https://huggingface.co/docs/optimum/onnxruntime/usage_guides/amdgpu

I am getting warning
```
1J_Li2ELi8ELi8ELb1ES1I_S1K_S1K_Li1ELi4ELi2ELb0ELi1ELi8ES1L_Li8ELNS1_21MaskingSpecializationE1ELi4ELS1M_0EE28ComputeBasePtrOfStridedBatchENS1_17C0Mat
rixMask_implINS1_29MaskOutUpperTrianglePredicateEEELb1EEEvPKT0_S2E_S2E_PT1_T2_T3_T4_T5_T6_T7_T8_T9_T10_T11_T12_T13_iT14_T15_': desired occupancy was
2, final occupancy is 1 [-Wpass-failed]
/app/libs/onnxruntime/build/Linux/Release/_deps/composable_kernel-src/include/ck/tensor_operation/gpu/device/impl/device_batched_gemm_softmax_gemm_p
ermute_xdl_cshuffle.hpp:47:9: warning: failed to meet occupancy target given by 'amdgpu-waves-per-eu' in '_ZN2ck16tensor_operation6device48kernel_ba
tched_gemm_softmax_gemm_xdl_cshuffle_v1INS_52GridwiseBatchedGemmMultipleDSoftmaxGemm_Xdl_CShuffleIDF16_fDF16_DF16_NS_5TupleIJEEENS0_12element_wise11
Pass ThroughES7 N11onnxruntime7contrib4rocm8internal26PreSoftmaxAttentionScoreOpES7 S7 LNS 25InMemoryDataOperationEnumEOENS 16TensorDescriptorINS4 IJ
```

Will this affect the performance of onnxruntime model?

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.