Merge branch 'develop' into aosewski/wavelet_omniperf

example/44_conv2d_fwd_quant/conv2d_fwd_xdl_perlayer_quantization_int8.cpp → example/44_conv2d_fwd_quantization/conv2d_fwd_xdl_perlayer_quantization_int8.cpp

@@ -11,6 +11,7 @@
 #include "ck/library/utility/device_memory.hpp"
 #include "ck/library/utility/host_tensor.hpp"
 #include "ck/library/utility/host_tensor_generator.hpp"
+#include "ck/library/utility/literals.hpp"
 #include "ck/library/utility/convolution_parameter.hpp"
 #include "ck/library/utility/convolution_host_tensor_descriptor_helper.hpp"
 #include "ck/library/reference_tensor_operation/cpu/reference_conv_fwd.hpp"
@@ -150,14 +151,14 @@ bool run_grouped_conv_fwd(bool do_verification,
     auto invoker  = conv.MakeInvoker();
     auto argument = conv.MakeArgument(in_device_buf.GetDeviceBuffer(),
                                       wei_device_buf.GetDeviceBuffer(),
-                                      std::array<const void*, 0>{},
+                                      {},
                                       out_device_buf.GetDeviceBuffer(),
                                       a_g_n_c_wis_lengths,
                                       a_g_n_c_wis_strides,
                                       b_g_k_c_xs_lengths,
                                       b_g_k_c_xs_strides,
-                                      std::array<std::array<ck::index_t, NDimSpatial + 3>, 0>{{}},
-                                      std::array<std::array<ck::index_t, NDimSpatial + 3>, 0>{{}},
+                                      {},
+                                      {},
                                       e_g_n_k_wos_lengths,
                                       e_g_n_k_wos_strides,
                                       conv_filter_strides,
@@ -213,8 +214,8 @@ bool run_grouped_conv_fwd(bool do_verification,
 
         out_device_buf.FromDevice(out_device.mData.data());
 
-        pass &= ck::utils::check_err(
-            out_device.mData, out_host.mData, "Error: incorrect results!", 1e-5f, 1e-4f);
+        pass &=
+            ck::utils::check_err(out_device, out_host, "Error: incorrect results!", 1e-5f, 1e-4f);
     }
 
     return (pass ? 0 : 1);

example/44_elementwise_permute/CMakeLists.txt

 add_example_executable(example_elementwise_permute_4D_fp16 elementwise_permute_4D_fp16.cpp)
+add_example_executable(example_elementwise_permute_4D_fp16_2d elementwise_permute_4D_fp16_2d.cpp)

example/46_gemm_add_multiply/CMakeLists.txt

+add_example_executable(example_gemm_add_multiply_dl_fp16 gemm_add_multiply_dl_fp16.cpp)
+add_example_executable(example_gemm_add_multiply_xdl_fp16 gemm_add_multiply_xdl_fp16.cpp)

include/ck/tensor_operation/gpu/device/device_elementwise_normalization.hpp

@@ -32,7 +32,7 @@ struct DeviceElementwiseNormalization : public BaseOperator
                         const std::vector<index_t> betaStrides,
                         const std::vector<index_t> yStrides,
                         const std::vector<index_t> reduceDims,
-                        AccDataType epsilon,
+                        double epsilon,
                         const std::array<const void*, NumInput> in_dev_buffers,
                         const void* p_gamma,
                         const void* p_beta,