From d270b881fcb888de806aeb05a279b8ca451d6f57 Mon Sep 17 00:00:00 2001 From: Bartlomiej Wroblewski Date: Mon, 27 Nov 2023 11:31:14 +0100 Subject: [PATCH] Fix cluster length arrange order in fp16 GEMM example (#1055) [ROCm/composable_kernel commit: bfecc19352140d107130d518428a0c51595cbd0c] --- example/01_gemm/gemm_xdl_fp16.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/example/01_gemm/gemm_xdl_fp16.cpp b/example/01_gemm/gemm_xdl_fp16.cpp index ef9e64cc69..2338cdc9c1 100644 --- a/example/01_gemm/gemm_xdl_fp16.cpp +++ b/example/01_gemm/gemm_xdl_fp16.cpp @@ -30,7 +30,7 @@ using DeviceGemmInstance0 = ck::tensor_operation::device::DeviceGemmXdl // ######| Type| Type| Type| Type| | | | Elementwise| Elementwise| Elementwise|Spacialization| Size| Block| Block| Block| | XDL| XDL| Per| Per| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraM| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| SrcDstVectorDim| DstScalar| // ######| | | | | | | | Operation| Operation| Operation| | | | | | | | | Wave| Wave| Lengths_K0_M_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | | PerVector| // ######| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - < ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementOp, BElementOp, CElementOp, GemmDefault, 256, 256, 128, 4, 8, 32, 32, 4, 2, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, 7, 1>; + < ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout, AElementOp, BElementOp, CElementOp, GemmDefault, 256, 256, 128, 4, 8, 32, 32, 4, 2, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 8, true, 7, 1>; // // clang-format on // clang-format off