accelerate dropout (#9902)

* accelerate dropout

* accelerate dropout

* "fix the dropout test"

* "rerun ci"

* "fix ci"

* "rerun ci"

* "fix ci"

* "fix"

* "stage"

* disable

paddle/fluid/operators/dropout_op.cu

@@ -24,21 +24,11 @@ namespace paddle {
 namespace operators {
 
 template <typename T>
-__global__ void RandomGenerator(const size_t n, const int seed,
-                                const float dropout_prob, const T* src,
-                                T* mask_data, T* dst) {
-  thrust::minstd_rand rng;
-  rng.seed(seed);
-  thrust::uniform_real_distribution<float> dist(0, 1);
-
+__global__ void RandomGenerator(const size_t n, const T* src,
+                                const T* cpu_mask_data, T* mask_data, T* dst) {
   int idx = blockDim.x * blockIdx.x + threadIdx.x;
   for (; idx < n; idx += blockDim.x * gridDim.x) {
-    rng.discard(idx);
-    if (dist(rng) < dropout_prob) {
-      mask_data[idx] = static_cast<T>(0);
-    } else {
-      mask_data[idx] = static_cast<T>(1);
-    }
+    mask_data[idx] = cpu_mask_data[idx];
     dst[idx] = mask_data[idx] * src[idx];
   }
 }
@@ -66,15 +56,27 @@ class GPUDropoutKernel : public framework::OpKernel<T> {
       std::random_device rnd;
       int seed =
           context.Attr<bool>("fix_seed") ? context.Attr<int>("seed") : rnd();
+      std::minstd_rand engine;
+      engine.seed(seed);
+      std::uniform_real_distribution<float> dist(0, 1);
+      framework::Vector<T> cpu_mask(size);
+      for (size_t i = 0; i < size; ++i) {
+        if (dist(engine) < dropout_prob) {
+          cpu_mask[i] = static_cast<T>(0);
+        } else {
+          cpu_mask[i] = static_cast<T>(1);
+        }
+      }
 
       int threads = 512;
       int grid = (x->numel() + threads - 1) / threads;
       RandomGenerator<
           T><<<grid, threads, 0, context.cuda_device_context().stream()>>>(
-          size, seed, dropout_prob, x_data, mask_data, y_data);
+          size, x_data, cpu_mask.CUDAData(context.GetPlace()), mask_data,
+          y_data);
     } else {
-      auto X = EigenMatrix<T>::Reshape(*x, 1);
-      auto Y = EigenMatrix<T>::Reshape(*y, 1);
+      auto X = EigenVector<T>::Flatten(*x);
+      auto Y = EigenVector<T>::Flatten(*y);
       Y.device(place) = X * static_cast<T>(1.0f - dropout_prob);
     }
   }
@@ -87,6 +89,8 @@ namespace ops = paddle::operators;
 namespace plat = paddle::platform;
 REGISTER_OP_CUDA_KERNEL(
     dropout, ops::GPUDropoutKernel<plat::CUDADeviceContext, float>,
+    ops::GPUDropoutKernel<plat::CUDADeviceContext, double>,
     ops::GPUDropoutKernel<plat::CUDADeviceContext, plat::float16>);
 REGISTER_OP_CUDA_KERNEL(dropout_grad,
+                        ops::DropoutGradKernel<plat::CUDADeviceContext, double>,
                         ops::DropoutGradKernel<plat::CUDADeviceContext, float>);

paddle/fluid/operators/dropout_op.h

@@ -24,7 +24,7 @@ namespace operators {
 using Tensor = framework::Tensor;
 template <typename T, int MajorType = Eigen::RowMajor,
           typename IndexType = Eigen::DenseIndex>
-using EigenMatrix = framework::EigenMatrix<T, MajorType, IndexType>;
+using EigenVector = framework::EigenVector<T, MajorType, IndexType>;
 
 template <typename DeviceContext, typename T>
 class CPUDropoutKernel : public framework::OpKernel<T> {
@@ -60,8 +60,8 @@ class CPUDropoutKernel : public framework::OpKernel<T> {
         }
       }
     } else {
-      auto X = EigenMatrix<T>::Reshape(*x, 1);
-      auto Y = EigenMatrix<T>::Reshape(*y, 1);
+      auto X = EigenVector<T>::Flatten(*x);
+      auto Y = EigenVector<T>::Flatten(*y);
       auto& place =
           *context.template device_context<DeviceContext>().eigen_device();
       Y.device(place) = X * (1.0f - dropout_prob);
@@ -81,9 +81,9 @@ class DropoutGradKernel : public framework::OpKernel<T> {
     auto* mask = context.Input<Tensor>("Mask");
     grad_x->mutable_data<T>(context.GetPlace());
 
-    auto M = EigenMatrix<T>::Reshape(*mask, 1);
-    auto dX = EigenMatrix<T>::Reshape(*grad_x, 1);
-    auto dY = EigenMatrix<T>::Reshape(*grad_y, 1);
+    auto M = EigenVector<T>::Flatten(*mask);
+    auto dX = EigenVector<T>::Flatten(*grad_x);
+    auto dY = EigenVector<T>::Flatten(*grad_y);
 
     auto& place =
         *context.template device_context<DeviceContext>().eigen_device();

paddle/fluid/operators/dropout_op_test.cc

@@ -13,6 +13,7 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #include <unistd.h>
+#include <iostream>
 
 #include <string>
 #include <thread>  // NOLINT
@@ -32,14 +33,16 @@ namespace m = paddle::operators::math;
 
 USE_OP(dropout);
 
+static paddle::framework::DDim dims = {10, 10};
+
 void Compare(f::Scope* scope, const p::DeviceContext& ctx) {
   // init
   auto var = scope->Var("X");
   auto tensor = var->GetMutable<f::LoDTensor>();
-  tensor->Resize({10, 10});
+  tensor->Resize(dims);
 
   std::vector<float> init;
-  for (int64_t i = 0; i < 10 * 10; ++i) {
+  for (int64_t i = 0; i < f::product(dims); ++i) {
     init.push_back(1.0);
   }
 
@@ -48,18 +51,19 @@ void Compare(f::Scope* scope, const p::DeviceContext& ctx) {
   auto place = ctx.GetPlace();
   auto out_var = scope->Var("Out");
   auto out_tensor = out_var->GetMutable<f::LoDTensor>();
-  out_tensor->Resize({10, 10});
+  out_tensor->Resize(dims);
   out_tensor->mutable_data<float>(place);  // allocate
 
   auto mask_var = scope->Var("Mask");
   auto mask_tensor = mask_var->GetMutable<f::LoDTensor>();
-  mask_tensor->Resize({10, 10});
+  mask_tensor->Resize(dims);
   mask_tensor->mutable_data<float>(place);  // allocate
 
   // run
   f::AttributeMap attrs;
   float dropout_prob = 0.5;
-  attrs.insert({"fix_seed", 1});
+  attrs.insert({"is_test", false});
+  attrs.insert({"fix_seed", true});
   attrs.insert({"seed", 3});
   attrs.insert({"dropout_prob", dropout_prob});
   auto dropout_op = f::OpRegistry::CreateOp(
@@ -69,6 +73,7 @@ void Compare(f::Scope* scope, const p::DeviceContext& ctx) {
 
   std::vector<float> out_vec;
   TensorToVector(*out_tensor, ctx, &out_vec);
+  ctx.Wait();
 
   std::vector<float> std_out = {
       0, 0, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1,
@@ -83,22 +88,22 @@ void Compare(f::Scope* scope, const p::DeviceContext& ctx) {
   }
 }
 
-// TODO(wyi): Due to
-// https://github.com/PaddlePaddle/Paddle/issues/9507, I temporarily
-// disable this test to remove the prevention of the merge of
-// unrelated PRs.
-/*
 TEST(Dropout, CPUDense) {
   f::Scope scope;
   p::CPUPlace place;
   p::CPUDeviceContext ctx(place);
-  Compare(scope, ctx);
+  Compare(&scope, ctx);
 }
 
+// TODO(wyi, dzhwinter): Due to
+// https://github.com/PaddlePaddle/Paddle/issues/9507, I temporarily
+// disable this test to remove the prevention of the merge of
+// unrelated PRs.
+/*
 TEST(Dropout, GPUDense) {
   f::Scope scope;
   p::CUDAPlace place;
   p::CUDADeviceContext ctx(place);
-  Compare(scope, ctx);
+  Compare(&scope, ctx);
 }
 */