Merge commit '57fadf6f' into gemm_layernorm_welford

421b718d · rocking · 2cf6f30b · 57fadf6f · 421b718d · 421b718d
Commit 421b718d authored Aug 26, 2022 by rocking
20 changed files
--- a/example/35_splitK_gemm/CMakeLists.txt
+++ b/example/35_splitK_gemm/CMakeLists.txt
--- a/example/35_splitK_gemm/run_splitK_gemm_example.inc
+++ b/example/35_splitK_gemm/run_splitK_gemm_example.inc
--- a/example/35_splitK_gemm/splitK_gemm_xdl_bfp16.cpp
+++ b/example/35_splitK_gemm/splitK_gemm_xdl_bfp16.cpp
--- a/example/35_splitK_gemm/splitK_gemm_xdl_fp16.cpp
+++ b/example/35_splitK_gemm/splitK_gemm_xdl_fp16.cpp
--- a/example/35_splitK_gemm/splitK_gemm_xdl_fp32.cpp
+++ b/example/35_splitK_gemm/splitK_gemm_xdl_fp32.cpp
--- a/example/35_splitK_gemm/splitK_gemm_xdl_int4.cpp
+++ b/example/35_splitK_gemm/splitK_gemm_xdl_int4.cpp
--- a/example/35_splitK_gemm/splitK_gemm_xdl_int8.cpp
+++ b/example/35_splitK_gemm/splitK_gemm_xdl_int8.cpp
--- a/example/CMakeLists.txt
+++ b/example/CMakeLists.txt
--- a/include/ck/tensor_operation/gpu/device/device_batched_contraction_multiple_d_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_contraction_multiple_d_xdl_cshuffle.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_gemm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_gemm_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_gemm_xdl_cshuffle.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm_permute.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm_permute.hpp
--- a/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm_permute_xdl_cshuffle.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d_multiple_r_xdl_cshuffle.hpp
--- a/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
--- a/include/ck/tensor_operation/gpu/device/gemm_specialization.hpp
+++ b/include/ck/tensor_operation/gpu/device/gemm_specialization.hpp