ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-20 14:59:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK] Fix misc issues in CK examples (#2890)

Browse Source 
* [CK] Fix misc CK issues

* revert fp8 change, it causes CI fail.

* resubmit fp8 change
This commit is contained in:
linqunAMD
2025-09-25 02:28:20 +08:00
committed by GitHub
parent 8fe3838c65
commit f076f207ce
10 changed files with 74 additions and 79 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
example/21_gemm_layernorm/gemm_bias_relu_add_layernorm_xdl_welford_fp16.cpp
View File

@@ -65,7 +65,7 @@ using DeviceOpInstance = ck::tensor_operation::device::DeviceGemmMultipleDLayern
//######| | | | | Type| Type| Type| DataType| Type| Type| Type| Type| Type| Elementwise| Elementwise| Elementwise| Elementwise| Spacialization| Prefetch| Size| Block| Block| Block| | | XDL| XDL| Per| Per| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraM| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| ThreadClusterLengths| ScalarPerVector| ThreadClusterLengths| ThreadSliceSize|
//######| | | | | | | | | | | | | | Operation| Operation| Operation| Operation| | Stage| | | | | | | | | Wave| Wave| Lengths_K0_M_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | PerShuffle| PerShuffle| _M_N| _M_N| _M_N| _M|
//######| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
< ALayout, BLayout, DsLayout, HLayout, ADataType, BDataType, AccDataType, CShuffleDataType, DsDataType, EMeanVarDataType, GammaDataType, BetaDataType, HDataType, AElementOp, BElementOp, CDEElementOp, HElementOp, GemmDefault, 1, 256, 256, 128, 32, 8, 8, 32, 32, 4, 2, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, 1, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, 1, 1, 1, S<32, 8>, 8, S<8, 32>, 8>;
< ALayout, BLayout, DsLayout, HLayout, ADataType, BDataType, AccDataType, CShuffleDataType, DsDataType, EMeanVarDataType, GammaDataType, BetaDataType, HDataType, AElementOp, BElementOp, CDEElementOp, HElementOp, GemmDefault, 1, 256, 256, 128, 32, 8, 8, 16, 16, 8, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, 1, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, 1, 1, 1, S<32, 8>, 4, S<8, 32>, 4>;
// clang-format on
auto f_host_tensor_descriptor1d = [](std::size_t len, std::size_t stride) {
@@ -154,8 +154,8 @@ void host_gemm_layernorm(Tensor<HDataType>& h_m_n,
int main()
{
// temp disable on gfx11 & gfx12
if(ck::is_gfx11_supported() || ck::is_gfx12_supported())
// temp disable on gfx11
if(ck::is_gfx11_supported())
{
return 0;
}