Skip to content

GitLab

"git@developer.sourcefind.cn:OpenDAS/fastmoe.git" did not exist on "2bcf98d70610bdff1d6f39f752949f6a0f8c5b89"
Commit 7bc4254d authored by ltqin's avatar ltqin
Browse files

fix vector load issue

parent 0e08b3b3
Hide whitespace changes
Inline Side-by-side
Showing with 3 additions and 3 deletions
host/driver_offline/include/device_convolution_backward_weight_implicit_gemm_v4r4r2_xdlops_nchw_kcyx_nkhw.hpp
View file @ 7bc4254d
...@@ -64,9 +64,9 @@ void device_convolution_backward_weight_implicit_gemm_v4r4r2_xdlops_nchw_kcyx_nk ...@@ -64,9 +64,9 @@ void device_convolution_backward_weight_implicit_gemm_v4r4r2_xdlops_nchw_kcyx_nk
using GemmABlockTransferThreadSliceLengths_GemmK0_GemmM_GemmK1 = Sequence<1, 2, 8>; using GemmABlockTransferThreadSliceLengths_GemmK0_GemmM_GemmK1 = Sequence<1, 2, 8>;
using GemmABlockTransferThreadClusterLengths_GemmK0_GemmM_GemmK1 = Sequence<4, 64, 1>; using GemmABlockTransferThreadClusterLengths_GemmK0_GemmM_GemmK1 = Sequence<4, 64, 1>;
// using vector load 4, so config's wo*ho must be a multiple of 4
constexpr index_t GemmABlockTransferSrcScalarPerVector_GemmK1 = 8; constexpr index_t GemmABlockTransferSrcScalarPerVector_GemmK1 = 4;
constexpr index_t GemmABlockTransferDstScalarPerVector_GemmK1 = 8; constexpr index_t GemmABlockTransferDstScalarPerVector_GemmK1 = 4;
using GemmBBlockTransferThreadSliceLengths_GemmK0_GemmN_GemmK1 = Sequence<1, 2, 8>; using GemmBBlockTransferThreadSliceLengths_GemmK0_GemmN_GemmK1 = Sequence<1, 2, 8>;
using GemmBBlockTransferThreadClusterLengths_GemmK0_GemmN_GemmK1 = Sequence<4, 64, 1>; using GemmBBlockTransferThreadClusterLengths_GemmK0_GemmN_GemmK1 = Sequence<4, 64, 1>;
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment