Merge branch 'develop' into gemm_layernorm_welford

5aa3c344 · rocking5566 · GitHub · 7fefc966 · 9d8f834a · 5aa3c344
Unverified Commit 5aa3c344 authored Oct 05, 2022 by rocking5566 Committed by GitHub Oct 05, 2022
20 changed files
--- a/include/ck/utility/span.hpp
+++ b/include/ck/utility/span.hpp
--- a/include/ck/utility/transpose_vectors.hpp
+++ b/include/ck/utility/transpose_vectors.hpp
--- a/library/include/ck/library/reference_tensor_operation/cpu/reference_groupnorm.hpp
+++ b/library/include/ck/library/reference_tensor_operation/cpu/reference_groupnorm.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute.hpp
--- a/library/include/ck/library/tensor_operation_instance/gpu/layernorm.hpp
+++ b/library/include/ck/library/tensor_operation_instance/gpu/layernorm.hpp
--- a/library/include/ck/library/utility/check_err.hpp
+++ b/library/include/ck/library/utility/check_err.hpp
--- a/library/include/ck/library/utility/fill.hpp
+++ b/library/include/ck/library/utility/fill.hpp
--- a/library/include/ck/library/utility/host_tensor.hpp
+++ b/library/include/ck/library/utility/host_tensor.hpp
--- a/library/src/tensor_operation_instance/gpu/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/CMakeLists.txt
+add_instance_library(device_batched_gemm_add_relu_gemm_add_instance
+    device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
+    device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gon_gmo_instance.cpp
+)
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gon_gmo_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_add_relu_gemm_add/device_batched_gemm_add_relu_gemm_add_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gon_gmo_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute/device_batched_gemm_masking_scale_softmax_gemm_permute_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_masking_scale_softmax_gemm_permute/device_batched_gemm_masking_scale_softmax_gemm_permute_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm/device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm/device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/gemm_bilinear/device_gemm_bilinear_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_bilinear/device_gemm_bilinear_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/normalization/device_layernorm_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/device_layernorm_f16_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/normalization/device_layernorm_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/device_layernorm_f32_instance.cpp
--- a/profiler/CMakeLists.txt
+++ b/profiler/CMakeLists.txt