aboutsummaryrefslogtreecommitdiff
path: root/src/cpu/kernels/CpuAddMulAddKernel.h
diff options
context:
space:
mode:
authorGunes Bayir <gunes.bayir@arm.com>2023-01-29 13:24:24 +0000
committerGunes Bayir <gunes.bayir@arm.com>2023-02-01 09:59:30 +0000
commitae72a46e495742863dba44fcf5fdc673c9d2afbc (patch)
tree65bab43d0feddaa66b160ac7dc746651dc7c48de /src/cpu/kernels/CpuAddMulAddKernel.h
parentec320d9fc418e2d95a3a38ce87233397535f467d (diff)
downloadComputeLibrary-ae72a46e495742863dba44fcf5fdc673c9d2afbc.tar.gz
Add new operator AddMulAdd for Neon™ backend for Float/Quantized types
This is a fused operator that merges Add + Mul + Add [+ Relu-based-Activation] layers and have an intermediate output after the first Add. It's supported for FP16/32/QASYMM8/QASYMM8_SIGNED data types. The subsequent Add and Mul are intended for scaling and the coefficients only have one dimension (per channel). The inputs are - input1 : nD tensor [X, Y, Z, W, ..] - input2 : nD tensor [X, Y, Z, W, ..] - add_coef : 1D tensor [X] - mul_coef : 1D tensor [X] The outputs are - out1 : nD tensor (intermediate output) [X, Y, Z, W, ..] - out2 : nD tensor (final output) [X, Y, Z, W, ..] The operation can be summarized as follows: out1 <- input1 + input2 out2 <- Act(out1 * mul_coef + add_coef) The activation function can be Identity, Relu, Bounded Relu or Lower/Upper Bounded Relu. The intermediate output can be skipped by providing a nullptr. The reason of providing this operator is to be able to fuse in case of Residual network patterns and save computations by reducing memory back and forward. Resolves: COMPMID-5463 Signed-off-by: Gunes Bayir <gunes.bayir@arm.com> Change-Id: I8ef577aa623b036e9a9f655cc088493fd19a6109 Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9055 Comments-Addressed: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Jakub Sujak <jakub.sujak@arm.com> Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com> Tested-by: Arm Jenkins <bsgcomp@arm.com> Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/cpu/kernels/CpuAddMulAddKernel.h')
-rw-r--r--src/cpu/kernels/CpuAddMulAddKernel.h89
1 files changed, 89 insertions, 0 deletions
diff --git a/src/cpu/kernels/CpuAddMulAddKernel.h b/src/cpu/kernels/CpuAddMulAddKernel.h
new file mode 100644
index 0000000000..595b580060
--- /dev/null
+++ b/src/cpu/kernels/CpuAddMulAddKernel.h
@@ -0,0 +1,89 @@
+/*
+ * Copyright (c) 2023 Arm Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef SRC_CPU_KERNELS_CPUADDMULADDKERNEL
+#define SRC_CPU_KERNELS_CPUADDMULADDKERNEL
+
+#include "src/core/common/Macros.h"
+#include "src/cpu/ICpuKernel.h"
+
+namespace arm_compute
+{
+namespace cpu
+{
+namespace kernels
+{
+/** Interface for the kernel to perform addition between two tensors */
+class CpuAddMulAddKernel : public ICpuKernel<CpuAddMulAddKernel>
+{
+private:
+ using AddMulAddKernelPtr =
+ std::add_pointer<void(const ITensor *, const ITensor *, const ITensor *, const ITensor *, ITensor *, ITensor *, ConvertPolicy, const ActivationLayerInfo &, const Window &)>::type;
+
+public:
+ struct AddMulAddKernel
+ {
+ const char *name;
+ const DataTypeISASelectorPtr is_selected;
+ AddMulAddKernelPtr ukernel;
+ };
+
+ CpuAddMulAddKernel() = default;
+ ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(CpuAddMulAddKernel);
+ /** Initialize the kernel's inputs and outputs.
+ *
+ * Similar to @ref NEAddMulAdd::configure()
+ *
+ */
+ void configure(const ITensorInfo *input1, const ITensorInfo *input2,
+ const ITensorInfo *bn_mul, const ITensorInfo *bn_add,
+ ITensorInfo *add_output, ITensorInfo *final_output,
+ ConvertPolicy policy, const ActivationLayerInfo &act_info);
+ /** Static function to check if given info will lead to a valid configuration
+ *
+ * Similar to CpuAddMulAddKernel::configure()
+ *
+ * @return a status
+ */
+ static Status validate(const ITensorInfo *input1, const ITensorInfo *input2,
+ const ITensorInfo *bn_mul, const ITensorInfo *bn_add,
+ const ITensorInfo *add_output, const ITensorInfo *final_output,
+ ConvertPolicy policy, const ActivationLayerInfo &act_info);
+
+ // Inherited methods overridden:
+ void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override;
+ const char *name() const override;
+
+ static const std::vector<AddMulAddKernel> &get_available_kernels();
+
+private:
+ ConvertPolicy _policy{};
+ ActivationLayerInfo _act_info{};
+ AddMulAddKernelPtr _run_method{ nullptr };
+ std::string _name{};
+};
+} // namespace kernels
+} // namespace cpu
+} // namespace arm_compute
+#endif /* SRC_CPU_KERNELS_CPUADDMULADDKERNEL */