Support broadcast for Power OP

mindspore/lite/src/ops/ops.h

@@ -207,6 +207,7 @@ class Power : public Primitive {
  public:
   explicit Power(schema::Primitive *primitive) : Primitive(primitive) {}
   const schema::Power *GetAttribute() const { return this->primitive->value_as_Power(); }
+  int InferShape(std::vector<tensor::Tensor *> inputs_, std::vector<tensor::Tensor *> outputs_) override;
 };
 
 class Range : public Primitive {

mindspore/lite/src/ops/power.cc

@@ -0,0 +1,50 @@
+/**
+ * Copyright 2019-2020 Huawei Technologies Co., Ltd
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#include <numeric>
+#include "src/ops/ops.h"
+#include "include/errorcode.h"
+#include "utils/log_adapter.h"
+#include "src/ir/tensor.h"
+
+namespace mindspore::lite {
+int Power::InferShape(std::vector<tensor::Tensor *> inputs, std::vector<tensor::Tensor *> outputs) {
+  MS_ASSERT(this->primitive != nullptr);
+  auto x_tensor = inputs[0];
+  MS_ASSERT(x_tensor != nullptr);
+  auto exp_tensor = inputs[1];
+  MS_ASSERT(exp_tensor != nullptr);
+  auto output_tensor = outputs[0];
+  MS_ASSERT(output_tensor != nullptr);
+  if (inputs.size() < 2) {
+    MS_LOG(ERROR) << "input size" << inputs.size() << " is error!";
+    return RET_INPUT_TENSOR_ERROR;
+  }
+  if (exp_tensor->shape() != x_tensor->shape() && exp_tensor->shape().size() != 1) {
+    MS_LOG(ERROR) << "Power inputs shape is not equal!";
+    return RET_INPUT_TENSOR_ERROR;
+  }
+
+  int exp_size = std::accumulate(exp_tensor->shape().begin(), exp_tensor->shape().end(), 1, std::multiplies<int>());
+  if (x_tensor->data_type() != exp_tensor->data_type() && exp_size != 1) {
+    MS_LOG(ERROR) << "Exponent tensor's shape is wrong";
+    return RET_INPUT_TENSOR_ERROR;
+  }
+  output_tensor->SetFormat(x_tensor->GetFormat());
+  output_tensor->set_shape(x_tensor->shape());
+  output_tensor->set_data_type(x_tensor->data_type());
+  return RET_OK;
+}
+}  // namespace mindspore::lite

mindspore/lite/src/runtime/kernel/arm/fp32/power.cc

@@ -50,13 +50,20 @@ int PowerCPUKernel::Run() {
 }
 
 int PowerCPUKernel::RunImpl(int task_id) {
-  auto input_addr = reinterpret_cast<float *>(inputs_.at(0)->Data());
+  auto x_addr = reinterpret_cast<float *>(inputs_[0]->Data());
-  auto output_addr = reinterpret_cast<float *>(outputs_.at(0)->Data());
+  auto exp_addr = reinterpret_cast<float *>(inputs_[1]->Data());
-  auto size = inputs_.at(0)->Size();
+  auto output_addr = reinterpret_cast<float *>(outputs_[0]->Data());
+  auto size = inputs_[0]->ElementsNum();
   int stride = UP_DIV(size, thread_count_);
   int len = MSMIN(stride, size - stride * task_id);
-
+  bool broadcast = (inputs_[1]->ElementsNum() == 1) ? true : false;
-  Power(input_addr + stride * task_id, output_addr + stride * task_id, len, power_, scale_, shift_);
+  float *cur_exp;
+  if (broadcast) {
+    cur_exp = exp_addr;
+  } else {
+    cur_exp = exp_addr + stride * task_id;
+  }
+  Power(x_addr + stride * task_id, cur_exp, output_addr + stride * task_id, len, scale_, shift_, broadcast);
   return RET_OK;
 }
 
@@ -67,7 +74,7 @@ kernel::LiteKernel *CpuPowerFp32KernelCreator(const std::vector<lite::tensor::Te
   MS_ASSERT(opParameter != nullptr);
   MS_ASSERT(desc.type == schema::PrimitiveType_Power);
   auto *kernel =
-    new (std::nothrow) PowerCPUKernel(reinterpret_cast<PowerParameter *>(opParameter), inputs, outputs, ctx);
+    new (std::nothrow) PowerCPUKernel(opParameter, inputs, outputs, ctx);
   if (kernel == nullptr) {
     MS_LOG(ERROR) << "new PowerCPUKernel fail!";
     return nullptr;

mindspore/lite/src/runtime/kernel/arm/fp32/power.h

@@ -18,20 +18,20 @@
 #define MINDSPORE_LITE_SRC_RUNTIME_KERNEL_ARM_FP32_POWER_H_
 
 #include <vector>
+#include "include/context.h"
 #include "src/lite_kernel.h"
-
 #include "src/runtime/kernel/arm/opclib/power.h"
 
 namespace mindspore::kernel {
 class PowerCPUKernel : public LiteKernel {
  public:
-  PowerCPUKernel(PowerParameter *param, const std::vector<lite::tensor::Tensor *> &inputs,
+  PowerCPUKernel(OpParameter *param, const std::vector<lite::tensor::Tensor *> &inputs,
                  const std::vector<lite::tensor::Tensor *> &outputs, const lite::Context *ctx)
-      : LiteKernel(reinterpret_cast<OpParameter *>(param), inputs, outputs),
+      : LiteKernel(param, inputs, outputs),
+        ctx_(ctx),
         thread_count_(ctx->thread_num_),
-        power_(param->power_),
+        scale_(reinterpret_cast<PowerParameter *>(opParameter)->scale_),
-        scale_(param->scale_),
+        shift_(reinterpret_cast<PowerParameter *>(opParameter)->shift_) {}
-        shift_(param->shift_) {}
   ~PowerCPUKernel() override = default;
 
   int Init() override;
@@ -40,8 +40,8 @@ class PowerCPUKernel : public LiteKernel {
   int RunImpl(int task_id);
 
  private:
+  const lite::Context *ctx_;
   int thread_count_;
-  float power_;
   float scale_;
   float shift_;
 };

mindspore/lite/src/runtime/kernel/arm/fp32/power_grad.cc

@@ -36,7 +36,8 @@ int PowerGradCPUKernel::Run() {
   auto dx_addr = reinterpret_cast<float *>(outputs_.at(0)->Data());
   auto size = inputs_.at(0)->ElementsNum();
 
-  Power(x_addr, dx_addr, size, power_ - 1, scale_, shift_);
+  float exp = power_ - 1;
+  Power(x_addr, &exp, dx_addr, size, scale_, shift_, true);
   ElementMul(dx_addr, dy_addr, dx_addr, size);
   float scale = scale_ * power_;
   for (int i = 0; i < size; i++) {

mindspore/lite/src/runtime/kernel/arm/opclib/power.cc

@@ -0,0 +1,58 @@
+/**
+ * Copyright 2020 Huawei Technologies Co., Ltd
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include "src/runtime/kernel/arm/opclib/power.h"
+
+bool CheckInteger(float f) { return floorf(f) == f; }
+
+float OptimizedPowerImpl(float x, int exponent) {
+  int exp = abs(exponent);
+  float result = 1;
+  float iterator = x;
+  while (exp) {
+    if (exp % 2) {
+      result *= iterator;
+    }
+    iterator *= iterator;
+    exp = exp / 2;
+  }
+  return exponent >= 0 ? result : 1 / result;
+}
+
+float StdPowerImpl(float x, float exponent) { return pow(x, exponent); }
+
+void Power(const float *input, const float *exponent, float *output, int len, float scale, float shift,
+           bool broadcast) {
+  if (broadcast) {
+    if (CheckInteger(*exponent)) {
+      for (int i = 0; i < len; ++i) {
+        output[i] = OptimizedPowerImpl(scale * input[i] + shift, (int)(*exponent));
+      }
+    } else {
+      for (int i = 0; i < len; ++i) {
+        output[i] = StdPowerImpl(scale * input[i] + shift, *exponent);
+      }
+    }
+  } else {
+    for (int i = 0; i < len; ++i) {
+      if (CheckInteger(*exponent)) {
+        output[i] = OptimizedPowerImpl(scale * input[i] + shift, (int)exponent[i]);
+      } else {
+        output[i] = StdPowerImpl(scale * input[i] + shift, exponent[i]);
+      }
+    }
+  }
+}

mindspore/lite/src/runtime/kernel/arm/opclib/power.h

@@ -26,11 +26,6 @@ struct PowerParameter {
   float shift_;
 };
 
-inline void Power(const float *input_data, float *output_data, int len, float power, float scale, float shift) {
+void Power(const float *input, const float *exponent, float *output, int len, float scale, float shift, bool broadcast);
-  for (int i = 0; i < len; ++i) {
-    output_data[i] = pow((scale * input_data[i] + shift), power);
-  }
-}
 
 #endif  // MINDSPORE_LITE_SRC_RUNTIME_KERNEL_ARM_OPCLIB_POWER_H_
-

mindspore/lite/test/ut/src/runtime/kernel/arm/fp32/power_fp32_tests.cc

@@ -0,0 +1,100 @@
+/**
+ * Copyright 2020 Huawei Technologies Co., Ltd
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#include "mindspore/core/utils/log_adapter.h"
+#include "common/common_test.h"
+#include "mindspore/lite/src/runtime/kernel/arm/fp32/power.h"
+#include "src/kernel_registry.h"
+#include "src/lite_kernel.h"
+
+namespace mindspore {
+class TestPowerFp32 : public mindspore::Common {
+ public:
+  TestPowerFp32() {}
+};
+
+int PowerTestInit(std::vector<lite::tensor::Tensor *> *inputs_, std::vector<lite::tensor::Tensor *> *outputs_,
+                  float *a_ptr, float *b_ptr, std::vector<int> a_shape, std::vector<int> b_shape,
+                  std::vector<int> c_shape) {
+  auto in_t =
+    new lite::tensor::Tensor(kNumberTypeFloat, a_shape, schema::Format_NHWC, static_cast<schema::NodeType>(1));
+  in_t->MallocData();
+  memcpy(in_t->Data(), a_ptr, sizeof(float) * in_t->ElementsNum());
+  inputs_->push_back(in_t);
+
+  auto weight_t =
+    new lite::tensor::Tensor(kNumberTypeFloat, b_shape, schema::Format_NHWC, static_cast<schema::NodeType>(1));
+  weight_t->MallocData();
+  memcpy(weight_t->Data(), b_ptr, sizeof(float) * weight_t->ElementsNum());
+  inputs_->push_back(weight_t);
+
+  auto out_t =
+    new lite::tensor::Tensor(kNumberTypeFloat, c_shape, schema::Format_NHWC, static_cast<schema::NodeType>(1));
+  out_t->MallocData();
+  outputs_->push_back(out_t);
+
+  return out_t->ElementsNum();
+}
+
+TEST_F(TestPowerFp32, Simple) {
+  std::vector<lite::tensor::Tensor *> inputs_;
+  std::vector<lite::tensor::Tensor *> outputs_;
+  auto param = new PowerParameter();
+  param->scale_ = 1;
+  param->shift_ = 0;
+  float a[] = {1, 2, 3, 4};
+  float b[] = {5, 6, 7, 8};
+  std::vector<int> a_shape = {2, 2};
+  std::vector<int> b_shape = {2, 2};
+  std::vector<int> c_shape = {2, 2};
+  int total_size = PowerTestInit(&inputs_, &outputs_, a, b, a_shape, b_shape, c_shape);
+  auto ctx = new lite::Context;
+  ctx->thread_num_ = 1;
+  auto op = new kernel::PowerCPUKernel(reinterpret_cast<OpParameter *>(param), inputs_, outputs_, ctx);
+  op->Init();
+  op->Run();
+  float correct[] = {1, 64, 2187, 65536};
+  float *output = reinterpret_cast<float *>(outputs_[0]->Data());
+  for (int i = 0; i < 4; ++i) printf("%f ", output[i]);
+  CompareOutputData(reinterpret_cast<float *>(outputs_[0]->Data()), correct, total_size, 0.0001);
+  delete op;
+  for (auto t : inputs_) delete t;
+  for (auto t : outputs_) delete t;
+}
+
+TEST_F(TestPowerFp32, Broadcast) {
+  std::vector<lite::tensor::Tensor *> inputs_;
+  std::vector<lite::tensor::Tensor *> outputs_;
+  auto param = new PowerParameter();
+  param->scale_ = 1;
+  param->shift_ = 0;
+  float a[] = {1, 2, 3, 4};
+  float b[] = {2};
+  std::vector<int> a_shape = {2, 2};
+  std::vector<int> b_shape = {1};
+  std::vector<int> c_shape = {2, 2};
+  int total_size = PowerTestInit(&inputs_, &outputs_, a, b, a_shape, b_shape, c_shape);
+  auto ctx = new lite::Context;
+  ctx->thread_num_ = 2;
+  auto op = new kernel::PowerCPUKernel(reinterpret_cast<OpParameter *>(param), inputs_, outputs_, ctx);
+  op->Init();
+  op->Run();
+  float correct[] = {1, 4, 9, 16};
+  CompareOutputData(reinterpret_cast<float *>(outputs_[0]->Data()), correct, total_size, 0.0001);
+  delete op;
+  for (auto t : inputs_) delete t;
+  for (auto t : outputs_) delete t;
+}
+}  // namespace mindspore