Skip to content

GitLab

"...git@developer.sourcefind.cn:OpenDAS/opencompass.git" did not exist on "2f1949e7a1ef908dad9454a88d21472f9ab8dbc7"
Commit 1014e6c9 authored by root's avatar root
Browse files

load a_block as whole

parent c23de07d
Expand all Show whitespace changes
Inline Side-by-side
Showing with 454 additions and 2 deletions
composable_kernel/include/driver/driver_dynamic_convolution_forward_implicit_gemm_v5r1_nchw_kcyx_nkhw.hpp
View file @ 1014e6c9
...@@ -165,7 +165,7 @@ struct DriverDynamicConvolutionForwardImplicitGemm_v5r1_nchw_kcyx_nkhw_pad ...@@ -165,7 +165,7 @@ struct DriverDynamicConvolutionForwardImplicitGemm_v5r1_nchw_kcyx_nkhw_pad
#if 1 #if 1
// GEMM // GEMM
using gridwise_gemm = GridwiseDynamicGemm_km_kn_mn_v2< using gridwise_gemm = GridwiseDynamicGemm_km_kn_mn_v3<
BlockSize, BlockSize,
Float, Float,
AccFloat, AccFloat,
......
driver/include/device_dynamic_convolution_forward_implicit_gemm_v5r1_nchw_kcyx_nkhw.hpp
View file @ 1014e6c9
...@@ -78,7 +78,7 @@ void device_dynamic_convolution_forward_implicit_gemm_v5r1_nchw_kcyx_nkhw(InDesc ...@@ -78,7 +78,7 @@ void device_dynamic_convolution_forward_implicit_gemm_v5r1_nchw_kcyx_nkhw(InDesc
constexpr index_t KPerThread = 8; constexpr index_t KPerThread = 8;
constexpr index_t HPerThread = 1; constexpr index_t HPerThread = 1;
constexpr index_t WPerThread = 1; constexpr index_t WPerThread = 1;
constexpr index_t CYXPerThread = 1; constexpr index_t CYXPerThread = 4;
using GemmABlockTransferThreadSliceLengths_GemmK_GemmM = Sequence<1, 1>; using GemmABlockTransferThreadSliceLengths_GemmK_GemmM = Sequence<1, 1>;
using GemmABlockTransferThreadClusterLengths_GemmK_GemmM = Sequence<4, 16>; using GemmABlockTransferThreadClusterLengths_GemmK_GemmM = Sequence<4, 16>;
......
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