Merge branch 'develop' into standalone-layernorm

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f32_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -52,7 +52,7 @@ ADD_BLOCKWISE_INST_REF_BY_ID(float, float, float, 4, 0, 1, 4, 1);
 ADD_BLOCKWISE_INST_REF_BY_ID(float, float, float, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f32_f64_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -28,7 +28,7 @@ ADD_BLOCKWISE_INST_REF_BY_ID(float, double, float, 7, 0, 0, 4, 1);
 ADD_BLOCKWISE_INST_REF_BY_ID(float, double, float, 7, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_f64_f64_f64.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -52,7 +52,7 @@ ADD_BLOCKWISE_INST_REF_BY_ID(double, double, double, 4, 0, 1, 4, 1);
 ADD_BLOCKWISE_INST_REF_BY_ID(double, double, double, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_i8_i32_i8.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_BLOCKWISE_INST_REF_BY_ID(int8_t, int32_t, int8_t, 5, 0, 0, 4, 1);
 ADD_BLOCKWISE_INST_REF_BY_ID(int8_t, int32_t, int8_t, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_blockwise_i8_i8_i8.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -40,7 +40,7 @@ ADD_BLOCKWISE_INST_REF_BY_ID(int8_t, int8_t, int8_t, 4, 0, 1, 4, 1);
 ADD_BLOCKWISE_INST_REF_BY_ID(int8_t, int8_t, int8_t, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_impl_common.hpp

@@ -6,7 +6,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 template <int BlockSize, int MThreadClusterSize, int KThreadClusterSize>
 struct ReductionConfiguration_1
@@ -34,7 +34,7 @@ struct ReductionConfiguration_2
 
 #define QUICK_REDUCE_TEST 1
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add.hpp

@@ -11,7 +11,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 using reduce_configuration_1_instances_multiblock_atomic_add = std::tuple<
     // clang-format off
@@ -193,7 +193,7 @@ void add_device_reduce_instance_multiblock_atomic_add(
                                                Rank,                                    \
                                                NumReduceDim)
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add_b16_f32_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(bhalf_t, float, float, 5, 0, 0, 4, 1);
 ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(bhalf_t, float, float, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add_f16_f32_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(half_t, float, float, 5, 0, 0, 4, 1);
 ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(half_t, float, float, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add_f32_f32_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(float, float, float, 5, 0, 0, 4, 1);
 ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(float, float, float, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add_f32_f64_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(float, double, float, 5, 0, 0, 4, 1);
 ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(float, double, float, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_multiblock_atomic_add_f64_f64_f64.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(double, double, double, 5, 0, 0, 4, 1);
 ADD_MULTIBLOCK_ATOMIC_ADD_INST_REF_BY_ID(double, double, double, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 #ifdef QUICK_REDUCE_TEST
 using reduce_configuration_2_instances_threadwise = std::tuple<
@@ -151,7 +151,7 @@ void add_device_reduce_instance_threadwise(
                                     Rank,                                    \
                                     NumReduceDim)
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_b16_f32_b16.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -53,7 +53,7 @@ ADD_THREADWISE_INST_REF_BY_ID(bhalf_t, float, bhalf_t, 4, 0, 1, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(bhalf_t, float, bhalf_t, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_f16_f16_f16.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -40,7 +40,7 @@ ADD_THREADWISE_INST_REF_BY_ID(half_t, half_t, half_t, 4, 0, 1, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(half_t, half_t, half_t, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_f16_f32_f16.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -28,7 +28,7 @@ ADD_THREADWISE_INST_REF_BY_ID(half_t, float, half_t, 7, 0, 0, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(half_t, float, half_t, 7, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_f32_f32_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -52,7 +52,7 @@ ADD_THREADWISE_INST_REF_BY_ID(float, float, float, 4, 0, 1, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(float, float, float, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_f32_f64_f32.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -28,7 +28,7 @@ ADD_THREADWISE_INST_REF_BY_ID(float, double, float, 7, 0, 0, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(float, double, float, 7, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_f64_f64_f64.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -52,7 +52,7 @@ ADD_THREADWISE_INST_REF_BY_ID(double, double, double, 4, 0, 1, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(double, double, double, 4, 0, 1, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck

library/include/ck/library/tensor_operation_instance/gpu/reduce/device_reduce_instance_threadwise_i8_i32_i8.hpp

@@ -10,7 +10,7 @@
 namespace ck {
 namespace tensor_operation {
 namespace device {
-namespace device_reduce_instance {
+namespace instance {
 
 // clang-format off
 // InDataType | AccDataType | OutDataType | ReduceOpId | NanPropaOpt | IndicesOpt | Rank | NumReduceDim 
@@ -24,7 +24,7 @@ ADD_THREADWISE_INST_REF_BY_ID(int8_t, int32_t, int8_t, 5, 0, 0, 4, 1);
 ADD_THREADWISE_INST_REF_BY_ID(int8_t, int32_t, int8_t, 5, 0, 0, 2, 1);
 // clang-format on
 
-} // namespace device_reduce_instance
+} // namespace instance
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck