!502 Gpu Support Gelu & GeluGrad

Merge pull request !502 from chenweifeng/gelu

mindspore/ccsrc/kernel/gpu/cuda_impl/gelu_impl.cu

@@ -0,0 +1,65 @@
+/**
+ * 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 "kernel/gpu/cuda_impl/gelu_impl.cuh"
+#include "device/gpu/cuda_common.h"
+
+template<typename T>
+__global__ void GeluKernel(size_t size, T* input_addr, T* output_addr) {
+  // formula:
+  // gelu(x) = 0.5 * x * (1.0 + tanh(y))
+  // tanh(y) = 2 / (1 + exp(-2y)) - 1)
+  // y = sqrt(2/pi) * (x + 0.044715 * x^3)
+  for (size_t pos = blockIdx.x * blockDim.x + threadIdx.x; pos < (size); pos += blockDim.x * gridDim.x) {
+    float x = input_addr[pos];
+    float tanh_res = tanh(0.7978845608 * (x + 0.044715 * x * x * x));
+    output_addr[pos] = 0.5 * x * (1.0 + tanh_res);
+  }
+}
+
+template<typename T>
+void Gelu(size_t size, T* input_addr, T* output_addr, cudaStream_t cuda_stream) {
+  GeluKernel<<<GET_BLOCKS(size), GET_THREADS, 0, cuda_stream>>>(size, input_addr, output_addr);
+  return;
+}
+
+
+template<typename T>
+__global__ void GeluGradKernel(size_t size, T* dy_addr, T* x_addr, T* dx_addr) {
+  // formula:
+  // dx = dy * y'
+  // y' = 0.5 * (1 + tanh(tanh_para)) +
+  //      0.5 * x * (1 - tanh(tanh_para) * tanh(tanh_para)) * mul_right
+  // tanh_para = sqrt(2/pi) * (x + 0.044715 * x^3)
+  // mul_right = sqrt(2/pi) * (1 + 3 * 0.044715 * x^2))
+  for (size_t pos = blockIdx.x * blockDim.x + threadIdx.x; pos < (size); pos += blockDim.x * gridDim.x) {
+    T x = x_addr[pos];
+    T tanh_res = tanh(0.7978845608  * (x + 0.044715 * x * x * x));
+    T mul_right = 0.7978845608  + 0.1070322244 * x * x;
+    T y_res = 0.5 * (1 + tanh_res) + 0.5 * x * (1 - tanh_res * tanh_res) * mul_right;
+    dx_addr[pos] = dy_addr[pos] * y_res;
+  }
+}
+
+template<typename T>
+void GeluGradKernel(size_t size, T* dy_addr, T* x_addr, T* dx_addr, cudaStream_t cuda_stream) {
+  GeluGradKernel<<<GET_BLOCKS(size), GET_THREADS, 0, cuda_stream>>>(size, dy_addr, x_addr, dx_addr);
+}
+
+
+template void Gelu(size_t size, float* input_addr, float* output_addr, cudaStream_t cuda_stream);
+template void GeluGradKernel(size_t size, float* dy_addr, float* x_addr, float* dx_addr, cudaStream_t cuda_stream);

mindspore/ccsrc/kernel/gpu/cuda_impl/gelu_impl.cuh

@@ -0,0 +1,27 @@
+/**
+ * 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_KERNEL_GPU_CUDA_IMP_GELU_H_
+#define MINDSPORE_CCSRC_KERNEL_GPU_CUDA_IMP_GELU_H_
+
+#include "device/gpu/cuda_common.h"
+template<typename T>
+void Gelu(size_t input_size, T* input_addr, T* output_addr, cudaStream_t cuda_stream);
+
+template<typename T>
+void GeluGradKernel(size_t size, T* dy_addr, T* x_addr, T* dx_addr, cudaStream_t cuda_stream);
+
+#endif  // MINDSPORE_CCSRC_KERNEL_GPU_CUDA_IMP_GELU_H_

mindspore/ccsrc/kernel/gpu/nn/gelu_grad_kernel.cc

@@ -0,0 +1,29 @@
+/**
+ * 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 "kernel/gpu/nn/gelu_grad_kernel.h"
+
+namespace mindspore {
+namespace kernel {
+MS_REG_GPU_KERNEL_ONE(GeluGrad,
+                      KernelAttr()
+                        .AddInputAttr(kNumberTypeFloat32)
+                        .AddInputAttr(kNumberTypeFloat32)
+                        .AddInputAttr(kNumberTypeFloat32)
+                        .AddOutputAttr(kNumberTypeFloat32),
+                      GeLUGpuGradKernel, float)
+}  // namespace kernel
+}  // namespace mindspore

mindspore/ccsrc/kernel/gpu/nn/gelu_grad_kernel.h

@@ -0,0 +1,75 @@
+/**
+ * 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_KERNEL_GPU_NN_GELU_GRAD_KERNEL_H_
+#define MINDSPORE_CCSRC_KERNEL_GPU_NN_GELU_GRAD_KERNEL_H_
+
+#include <vector>
+#include "kernel/gpu/gpu_kernel.h"
+#include "kernel/gpu/gpu_kernel_factory.h"
+#include "kernel/gpu/kernel_constants.h"
+#include "kernel/gpu/cuda_impl/gelu_impl.cuh"
+
+namespace mindspore {
+namespace kernel {
+template <typename T>
+class GeLUGpuGradKernel : public GpuKernel {
+ public:
+  GeLUGpuGradKernel() : input_size_(0) {}
+  ~GeLUGpuGradKernel() override = default;
+  const std::vector<size_t> &GetInputSizeList() const override { return input_size_list_; }
+  const std::vector<size_t> &GetOutputSizeList() const override { return output_size_list_; }
+  const std::vector<size_t> &GetWorkspaceSizeList() const override { return workspace_size_list_; }
+
+  bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &,
+              const std::vector<AddressPtr> &outputs, uintptr_t stream_ptr) override {
+    T *dy_addr = GetDeviceAddress<T>(inputs, 0);
+    T *x_addr = GetDeviceAddress<T>(inputs, 1);
+    T *dx_addr = GetDeviceAddress<T>(outputs, 0);
+
+    GeluGradKernel(input_size_ / sizeof(T), dy_addr, x_addr, dx_addr, reinterpret_cast<cudaStream_t>(stream_ptr));
+    return true;
+  }
+
+  bool Init(const CNodePtr &kernel_node) override {
+    InitResource();
+    input_size_ = sizeof(T);
+    auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, 0);
+    for (auto dim : input_shape) {
+      input_size_ *= dim;
+    }
+    InitSizeLists();
+    return true;
+  }
+
+ protected:
+  void InitSizeLists() override {
+    input_size_list_.push_back(input_size_);
+    input_size_list_.push_back(input_size_);
+    input_size_list_.push_back(input_size_);
+    output_size_list_.push_back(input_size_);
+  }
+
+ private:
+  std::vector<size_t> input_size_list_;
+  std::vector<size_t> output_size_list_;
+  std::vector<size_t> workspace_size_list_;
+  size_t input_size_;
+};
+}  // namespace kernel
+}  // namespace mindspore
+
+#endif  // MINDSPORE_CCSRC_KERNEL_GPU_NN_GELU_GRAD_KERNEL_H_

mindspore/ccsrc/kernel/gpu/nn/gelu_kernel.cc

@@ -0,0 +1,24 @@
+/**
+ * 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 "kernel/gpu/nn/gelu_kernel.h"
+
+namespace mindspore {
+namespace kernel {
+MS_REG_GPU_KERNEL_ONE(Gelu, KernelAttr().AddInputAttr(kNumberTypeFloat32).AddOutputAttr(kNumberTypeFloat32),
+                      GeluGpuKernel, float)
+}  // namespace kernel
+}  // namespace mindspore

mindspore/ccsrc/kernel/gpu/nn/gelu_kernel.h

@@ -0,0 +1,72 @@
+/**
+ * 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_KERNEL_GPU_NN_GELU_GPU_KERNEL_H_
+#define MINDSPORE_CCSRC_KERNEL_GPU_NN_GELU_GPU_KERNEL_H_
+
+#include <vector>
+#include "kernel/gpu/gpu_kernel.h"
+#include "kernel/gpu/gpu_kernel_factory.h"
+#include "kernel/gpu/kernel_constants.h"
+#include "kernel/gpu/cuda_impl/gelu_impl.cuh"
+
+namespace mindspore {
+namespace kernel {
+template <typename T>
+class GeluGpuKernel : public GpuKernel {
+ public:
+  GeluGpuKernel() : input_size_(0) {}
+  ~GeluGpuKernel() override = default;
+  const std::vector<size_t> &GetInputSizeList() const override { return input_size_list_; }
+  const std::vector<size_t> &GetOutputSizeList() const override { return output_size_list_; }
+  const std::vector<size_t> &GetWorkspaceSizeList() const override { return workspace_size_list_; }
+
+  bool Launch(const std::vector<AddressPtr> &inputs, const std::vector<AddressPtr> &,
+              const std::vector<AddressPtr> &outputs, uintptr_t stream_ptr) override {
+    T *input_addr = GetDeviceAddress<T>(inputs, 0);
+    T *output_addr = GetDeviceAddress<T>(outputs, 0);
+
+    Gelu(input_size_ / sizeof(T), input_addr, output_addr, reinterpret_cast<cudaStream_t>(stream_ptr));
+    return true;
+  }
+
+  bool Init(const CNodePtr &kernel_node) override {
+    InitResource();
+    input_size_ = sizeof(T);
+    auto input_shape = AnfAlgo::GetPrevNodeOutputInferShape(kernel_node, 0);
+    for (auto dim : input_shape) {
+      input_size_ *= dim;
+    }
+    InitSizeLists();
+    return true;
+  }
+
+ protected:
+  void InitSizeLists() override {
+    input_size_list_.push_back(input_size_);
+    output_size_list_.push_back(input_size_);
+  }
+
+ private:
+  std::vector<size_t> input_size_list_;
+  std::vector<size_t> output_size_list_;
+  std::vector<size_t> workspace_size_list_;
+  size_t input_size_;
+};
+}  // namespace kernel
+}  // namespace mindspore
+
+#endif  // MINDSPORE_CCSRC_KERNEL_GPU_NN_GELU_GPU_KERNEL_H_

tests/st/ops/gpu/test_gelu_grad_op.py

@@ -0,0 +1,61 @@
+# 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 numpy as np
+from mindspore import Tensor
+from mindspore.ops import operations as P
+import mindspore.nn as nn
+import mindspore.context as context
+from mindspore.ops import composite as C
+
+context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+
+class GeluNet(nn.Cell):
+    def __init__(self):
+        super(GeluNet, self).__init__()
+        self.gelu = P.Gelu()
+
+    def construct(self, x):
+        return self.gelu(x)
+
+class Grad(nn.Cell):
+    def __init__(self, network):
+        super(Grad, self).__init__()
+        self.grad = C.GradOperation(name="get_all", get_all=True, sens_param=True)
+        self.network = network
+
+    def construct(self, input_data, sens):
+        gout = self.grad(self.network)(input_data, sens)
+        return gout
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gelugrad():
+    x_ms = Tensor(np.array([0.58401114, 0.68800163, 0.9760397,  0.14702141, 0.46563736, 0.9607501,
+                            0.14567593, 0.12261796, 0.37054458, 0.46421242]).astype(np.float32))
+    dy_ms = Tensor(np.array([0.5559598,  0.96994054, 0.24770357, 0.34646875, 0.2984393,  0.03287048,
+                             0.55681044, 0.966908,   0.06015943, 0.6099489 ]).astype(np.float32))
+
+    net = GeluNet()
+    grad = Grad(net)
+
+    output = grad(x_ms, dy_ms)
+    print(output)
+    expect = [0.50963277, 0.9414753,  0.2667653,  0.21358444, 0.25243032, 0.0352667,
+              0.34266686, 0.57757664, 0.04707306, 0.51536125]
+    assert np.allclose(output[0].asnumpy(), expect)

tests/st/ops/gpu/test_gelu_op.py

@@ -0,0 +1,88 @@
+# 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
+from mindspore import Tensor
+from mindspore.ops import operations as P
+import mindspore.nn as nn
+import numpy as np
+import mindspore.context as context
+
+context.set_context(mode=context.GRAPH_MODE, device_target="GPU")
+
+class GeluNet(nn.Cell):
+    def __init__(self):
+        super(GeluNet, self).__init__()
+        self.gelu = P.Gelu()
+
+    def construct(self, x):
+        return self.gelu(x)
+
+
+def GeluCompute(x):
+    return 0.5 * x * (1.0 + np.tanh(np.sqrt(2 / np.pi) * (x + 0.044715 * x * x * x)))
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gelu_1d():
+    x_np = np.random.random((50,)).astype(np.float32)
+    y_np = GeluCompute(x_np)
+
+    x_ms = Tensor(x_np)
+    net = GeluNet()
+    y_ms = net(x_ms)
+
+    assert np.allclose(y_np, y_ms.asnumpy())
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gelu_2d():
+    x_np = np.random.random((50, 40)).astype(np.float32)
+    y_np = GeluCompute(x_np)
+
+    x_ms = Tensor(x_np)
+    net = GeluNet()
+    y_ms = net(x_ms)
+
+    assert np.allclose(y_np, y_ms.asnumpy())
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gelu_4d():
+    x_np = np.random.random((32, 3, 224, 224)).astype(np.float32)
+    y_np = GeluCompute(x_np)
+
+    x_ms = Tensor(x_np)
+    net = GeluNet()
+    y_ms = net(x_ms)
+
+    assert np.allclose(y_np, y_ms.asnumpy())
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.env_onecard
+def test_gelu_neg():
+    x_np = np.random.random((32, 3, 224, 224)).astype(np.float32) * -1
+    y_np = GeluCompute(x_np)
+
+    x_ms = Tensor(x_np)
+    net = GeluNet()
+    y_ms = net(x_ms)
+
+    assert np.allclose(y_np, y_ms.asnumpy())