Change the example to f16

example/53_layernorm2d_bwd/CMakeLists.txt

-add_example_executable(example_layernorm2d_bwd_fp16 layernorm2d_bwd_fp16.cpp)
+add_example_executable(example_layernorm2d_bwd_fp32 layernorm2d_bwd_fp32.cpp)

example/53_layernorm2d_bwd/layernorm2d_bwd_fp16.cpp → example/53_layernorm2d_bwd/layernorm2d_bwd_fp32.cpp

@@ -19,13 +19,13 @@
 #include "ck/tensor_operation/gpu/device/impl/device_normalization_bwd_gamma_beta_impl.hpp"
 #include "ck/library/reference_tensor_operation/cpu/reference_layernorm_bwd.hpp"
 
-using DYDataType         = ck::half_t;
-using XDataType          = ck::half_t;
-using GammaDataType      = ck::half_t;
+using DYDataType         = float;
+using XDataType          = float;
+using GammaDataType      = float;
 using MeanInvStdDataType = float;
-using DGammaDataType     = ck::half_t;
-using DBetaDataType      = ck::half_t;
-using DXDataType         = ck::half_t;
+using DGammaDataType     = float;
+using DBetaDataType      = float;
+using DXDataType         = float;
 using ComputeDataType    = float;
 
 constexpr int Rank         = 2;
@@ -64,17 +64,17 @@ using XDeviceInstance = ck::tensor_operation::device::DeviceNormalizationBwdXImp
     8,     // MThreadClusterSize
     32,    // KThreadClusterSize
     1,     // MThreadSliceSize
-    8,     // KThreadSliceSize
+    4,     // KThreadSliceSize
     true,  // IsDYFastestDimReduced
-    8,     // DYSrcVectorSize
+    4,     // DYSrcVectorSize
     true,  // IsXFastestDimReduced
-    8,     // XSrcVectorSize
+    4,     // XSrcVectorSize
     true,  // IsGammaFastestDimReduced
-    8,     // GammaSrcVectorSize
+    4,     // GammaSrcVectorSize
     false, // IsMeanInvStdFastestDimReduced
     1,     // MeanInvStdSrcVectorSize
     true,  // IsDXFastestDimReduced
-    8>;    // DXDstVectorSize
+    4>;    // DXDstVectorSize
 
 using GammaBetaDeviceInstance = ck::tensor_operation::device::DeviceNormalizationBwdGammaBetaImpl<
     DYDataType,
@@ -88,16 +88,16 @@ using GammaBetaDeviceInstance = ck::tensor_operation::device::DeviceNormalizatio
     256,   // BlockSize
     8,     // MThreadClusterSize
     32,    // KThreadClusterSize
-    8,     // MThreadSliceSize
+    4,     // MThreadSliceSize
     1,     // KThreadSliceSize
     false, // IsDYFastestDimReduced
-    8,     // DYSrcVectorSize
+    4,     // DYSrcVectorSize
     false, // IsXFastestDimReduced
-    8,     // XSrcVectorSize
+    4,     // XSrcVectorSize
     true,  // IsMeanInvStdFastestDimReduced
     1,     // MeanInvStdSrcVectorSize
-    8,     // DGammaDstVectorSize
-    8>;    // DBetaDstVectorSize
+    4,     // DGammaDstVectorSize
+    4>;    // DBetaDstVectorSize
 
 int main()
 {

example/54_groupnorm_bwd/CMakeLists.txt

-add_example_executable(example_groupnorm_bwd_fp16 groupnorm_bwd_fp16.cpp)
+add_example_executable(example_groupnorm_bwd_fp32 groupnorm_bwd_fp32.cpp)

example/54_groupnorm_bwd/groupnorm_bwd_fp16.cpp → example/54_groupnorm_bwd/groupnorm_bwd_fp32.cpp

@@ -19,13 +19,13 @@
 #include "ck/tensor_operation/gpu/device/impl/device_normalization_bwd_gamma_beta_impl.hpp"
 #include "ck/library/reference_tensor_operation/cpu/reference_groupnorm_bwd.hpp"
 
-using DYDataType         = ck::half_t;
-using XDataType          = ck::half_t;
-using GammaDataType      = ck::half_t;
+using DYDataType         = float;
+using XDataType          = float;
+using GammaDataType      = float;
 using MeanInvStdDataType = float;
-using DGammaDataType     = ck::half_t;
-using DBetaDataType      = ck::half_t;
-using DXDataType         = ck::half_t;
+using DGammaDataType     = float;
+using DBetaDataType      = float;
+using DXDataType         = float;
 using ComputeDataType    = float;
 
 constexpr int Rank         = 5;
@@ -54,17 +54,17 @@ using XDeviceInstance = ck::tensor_operation::device::DeviceNormalizationBwdXImp
     8,     // MThreadClusterSize
     32,    // KThreadClusterSize
     1,     // MThreadSliceSize
-    8,     // KThreadSliceSize
+    4,     // KThreadSliceSize
     true,  // IsDYFastestDimReduced
-    8,     // DYSrcVectorSize
+    4,     // DYSrcVectorSize
     true,  // IsXFastestDimReduced
-    8,     // XSrcVectorSize
+    4,     // XSrcVectorSize
     true,  // IsGammaFastestDimReduced
-    8,     // GammaSrcVectorSize
+    4,     // GammaSrcVectorSize
     false, // IsMeanInvStdFastestDimReduced
     1,     // MeanInvStdSrcVectorSize
     true,  // IsDXFastestDimReduced
-    8>;    // DXDstVectorSize
+    4>;    // DXDstVectorSize
 
 // kernel 2:                  M    , K
 // dy:     N, H, W, G, C -> G * C, N * H * W
@@ -89,16 +89,16 @@ using GammaBetaDeviceInstance = ck::tensor_operation::device::DeviceNormalizatio
     256,   // BlockSize
     8,     // ClusterInvarient
     32,    // ClusterReduce
-    8,     // SliceInvarient
+    4,     // SliceInvarient
     1,     // SliceReduce
     false, // IsDYFastestDimReduced
-    8,     // DYSrcVectorSize
+    4,     // DYSrcVectorSize
     false, // IsXFastestDimReduced
-    8,     // XSrcVectorSize
+    4,     // XSrcVectorSize
     false, // IsMeanInvStdFastestDimReduced
     1,     // MeanInvStdSrcVectorSize
-    8,     // DGammaDstVectorSize
-    8>;    // DBetaDstVectorSize
+    4,     // DGammaDstVectorSize
+    4>;    // DBetaDstVectorSize
 
 int main()
 {