!9323 Add StandardNormal for CPU

From: @yuanwei66
Reviewed-by: 
Signed-off-by:

mindspore/ccsrc/backend/kernel_compiler/cpu/random_cpu_kernel.cc

@@ -0,0 +1,110 @@
+/**
+ * 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 <random>
+#include <thread>
+#include "runtime/device/cpu/cpu_device_address.h"
+#include "backend/kernel_compiler/cpu/random_cpu_kernel.h"
+
+namespace mindspore {
+namespace kernel {
+void StandardNormal(float *output, std::normal_distribution<float> distribution,
+                    std::default_random_engine random_generator, size_t start, size_t end) {
+  for (size_t i = start; i < end; i++) {
+    output[i] = distribution(random_generator);
+  }
+}
+void LaunchStandardNormal(int seed, int seed2, const std::vector<AddressPtr> &outputs) {
+  unsigned int RNG_seed;
+  std::random_device rd;
+  if (seed2 != 0) {
+    RNG_seed = IntToUint(seed2);
+  } else if (seed != 0) {
+    RNG_seed = IntToUint(seed);
+  } else {
+    RNG_seed = rd();
+  }
+
+  auto output = reinterpret_cast<float *>(outputs[0]->addr);
+  // multithreading
+  size_t lens = outputs[0]->size / sizeof(float);
+  auto max_thread_num = std::thread::hardware_concurrency();
+  size_t thread_num = lens < 128 * max_thread_num ? std::ceil(lens / 128.0) : max_thread_num;
+  if (thread_num < 1) {
+    MS_LOG(ERROR) << "Invalid value: thread_num " << thread_num;
+    return;
+  }
+  std::vector<std::thread> threads;
+  threads.reserve(thread_num);
+  size_t start = 0;
+  size_t once_compute_size = (lens + thread_num - 1) / thread_num;
+  if (once_compute_size < 1) {
+    MS_LOG(ERROR) << "Invalid value: once_compute_size " << once_compute_size;
+    return;
+  }
+  std::normal_distribution<float> distribution;
+  while (start < lens) {
+    // avoid different threads using the same seed to generate the same random number
+    std::default_random_engine random_generator(++RNG_seed);
+    size_t end = (start + once_compute_size) > lens ? lens : (start + once_compute_size);
+    threads.emplace_back(std::thread(StandardNormal, output, distribution, random_generator, start, end));
+    start += once_compute_size;
+  }
+  for (size_t i = 0; i < threads.size(); ++i) {
+    threads[i].join();
+  }
+}
+
+void RandomCPUKernel::InitKernel(const CNodePtr &kernel_node) {
+  MS_EXCEPTION_IF_NULL(kernel_node);
+  std::string kernel_name = AnfAlgo::GetCNodeName(kernel_node);
+  auto iter = kRandomOpTypeMap.find(kernel_name);
+  if (iter == kRandomOpTypeMap.end()) {
+    MS_LOG(EXCEPTION) << "Random operation " << kernel_name << " is not supported.";
+  } else {
+    random_op_type_ = iter->second;
+  }
+
+  size_t input_num = AnfAlgo::GetInputTensorNum(kernel_node);
+  if ((random_op_type_ == RANDOM_OP_NORMAL) && input_num != 1) {
+    MS_LOG(EXCEPTION) << "Input number is " << input_num << ", but random op needs 1 input.";
+  }
+
+  size_t output_num = AnfAlgo::GetOutputTensorNum(kernel_node);
+  if (output_num != 1) {
+    MS_LOG(EXCEPTION) << "Output number is " << output_num << ", but random op needs 1 output.";
+  }
+
+  seed_ = LongToInt(GetValue<int64_t>(AnfAlgo::GetCNodePrimitive(kernel_node)->GetAttr("seed")));
+  seed2_ = LongToInt(GetValue<int64_t>(AnfAlgo::GetCNodePrimitive(kernel_node)->GetAttr("seed2")));
+}
+
+bool RandomCPUKernel::Launch(const std::vector<kernel::AddressPtr> &inputs,
+                             const std::vector<kernel::AddressPtr> & /*workspace*/,
+                             const std::vector<kernel::AddressPtr> &outputs) {
+  switch (random_op_type_) {
+    case RANDOM_OP_NORMAL: {
+      LaunchStandardNormal(seed_, seed2_, outputs);
+      break;
+    }
+    default: {
+      MS_LOG(EXCEPTION) << "Random operation " << random_op_type_ << " is not supported.";
+    }
+  }
+  return true;
+}
+
+}  // namespace kernel
+}  // namespace mindspore

mindspore/ccsrc/backend/kernel_compiler/cpu/random_cpu_kernel.h

@@ -0,0 +1,50 @@
+/**
+ * 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.
+ */
+#ifndef MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_RANDOM_CPU_KERNEL_H_
+#define MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_RANDOM_CPU_KERNEL_H_
+#include <vector>
+#include <string>
+#include <map>
+#include "backend/kernel_compiler/cpu/cpu_kernel.h"
+#include "backend/kernel_compiler/cpu/cpu_kernel_factory.h"
+
+namespace mindspore {
+namespace kernel {
+enum RandomOptype { RANDOM_OP_NORMAL = 0, RANDOM_OP_UNIFORM_INT, RANDOM_OP_UNIFORM_REAL, RANDOM_OP_INVALID_TYPE = 255 };
+
+const std::map<std::string, RandomOptype> kRandomOpTypeMap = {
+  {"StandardNormal", RANDOM_OP_NORMAL}, {"UniformInt", RANDOM_OP_UNIFORM_INT}, {"UniformReal", RANDOM_OP_UNIFORM_REAL}};
+
+class RandomCPUKernel : public CPUKernel {
+ public:
+  RandomCPUKernel() = default;
+  ~RandomCPUKernel() override = default;
+  void InitKernel(const CNodePtr &kernel_node) override;
+
+  bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &workspace,
+              const std::vector<AddressPtr> &outputs) override;
+
+ private:
+  RandomOptype random_op_type_{RANDOM_OP_INVALID_TYPE};
+  int seed_{0};
+  int seed2_{0};
+};
+
+MS_REG_CPU_KERNEL(StandardNormal, KernelAttr().AddInputAttr(kNumberTypeInt32).AddOutputAttr(kNumberTypeFloat32),
+                  RandomCPUKernel);
+}  // namespace kernel
+}  // namespace mindspore
+#endif  // MINDSPORE_CCSRC_BACKEND_KERNEL_COMPILER_CPU_RANDOM_CPU_KERNEL_H_

mindspore/core/utils/convert_utils_base.h

@@ -98,6 +98,13 @@ inline uint64_t LongToUlong(int64_t u) {
   return static_cast<uint64_t>(u);
 }
 
+inline int32_t LongToInt(int64_t u) {
+  if (u > static_cast<int64_t>((std::numeric_limits<int32_t>::max)())) {
+    MS_LOG(EXCEPTION) << "The size_t value(" << u << ") exceeds the maximum value of int.";
+  }
+  return static_cast<int32_t>(u);
+}
+
 inline int64_t UlongToLong(uint64_t u) {
   if (u > static_cast<uint64_t>((std::numeric_limits<int64_t>::max)())) {
     MS_LOG(EXCEPTION) << "The uint64_t value(" << u << ") exceeds the maximum value of int64_t.";

mindspore/ops/operations/random_ops.py

@@ -35,7 +35,7 @@ class StandardNormal(PrimitiveWithInfer):
         Tensor. The shape is the same as the input `shape`. The dtype is float32.
 
     Supported Platforms:
-        ``Ascend`` ``GPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> shape = (4, 16)

tests/st/ops/cpu/test_standard_normal.py

@@ -0,0 +1,85 @@
+# 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.
+# ============================================================================
+
+import pytest
+import mindspore.context as context
+import mindspore.nn as nn
+from mindspore.ops import operations as P
+from scipy.stats import kstest
+
+context.set_context(mode=context.GRAPH_MODE, device_target="CPU")
+
+
+class Net(nn.Cell):
+    def __init__(self, shape, seed=0, seed2=0):
+        super(Net, self).__init__()
+        self.shape = shape
+        self.seed = seed
+        self.seed2 = seed2
+        self.stdnormal = P.StandardNormal(seed, seed2)
+
+    def construct(self):
+        return self.stdnormal(self.shape)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_net():
+    seed = 10
+    seed2 = 10
+    shape = (5, 6, 8)
+    net = Net(shape, seed, seed2)
+    output = net()
+    assert output.shape == (5, 6, 8)
+    outnumpyflatten_1 = output.asnumpy().flatten()
+    _, p_value = kstest(outnumpyflatten_1, "norm")
+    # p-value is greater than the  significance level, cannot reject the hypothesis that the data come from
+    # the standard norm distribution.
+    assert p_value >= 0.05
+
+    seed = 0
+    seed2 = 10
+    shape = (5, 6, 8)
+    net = Net(shape, seed, seed2)
+    output = net()
+    assert output.shape == (5, 6, 8)
+    outnumpyflatten_2 = output.asnumpy().flatten()
+    _, p_value = kstest(outnumpyflatten_2, "norm")
+    assert p_value >= 0.05
+    # same seed should generate same random number
+    assert (outnumpyflatten_1 == outnumpyflatten_2).all()
+
+    seed = 0
+    seed2 = 0
+    shape = (130, 120, 141)
+    net = Net(shape, seed, seed2)
+    output = net()
+    assert output.shape == (130, 120, 141)
+    outnumpyflatten_1 = output.asnumpy().flatten()
+    _, p_value = kstest(outnumpyflatten_1, "norm")
+    assert p_value >= 0.05
+
+    seed = 0
+    seed2 = 0
+    shape = (130, 120, 141)
+    net = Net(shape, seed, seed2)
+    output = net()
+    assert output.shape == (130, 120, 141)
+    outnumpyflatten_2 = output.asnumpy().flatten()
+    _, p_value = kstest(outnumpyflatten_2, "norm")
+    assert p_value >= 0.05
+    # different seed(seed = 0) should generate different random number
+    assert ~(outnumpyflatten_1 == outnumpyflatten_2).all()