!3966 [AutoParallel]Add dropout distributed op

Merge pull request !3966 from lichen/add_dropout_distributed_op

mindspore/ccsrc/frontend/parallel/dynamic_creator.h

@@ -135,6 +135,7 @@ REGISTER(GatherV2PInfo);
 REGISTER(EmbeddingLookupInfo);
 REGISTER(TileInfo);
 REGISTER(StridedSliceInfo);
+REGISTER(DropoutInfo);
 }  // namespace parallel
 }  // namespace mindspore
 

mindspore/ccsrc/frontend/parallel/ops_info/activation_info.cc

@@ -20,6 +20,8 @@
 #include <memory>
 #include <vector>
 #include <utility>
+#include <functional>
+#include <numeric>
 
 #include "ir/value.h"
 #include "frontend/parallel/auto_parallel/costmodel.h"
@@ -54,6 +56,29 @@ Status Activation::CheckStrategy(const StrategyPtr &strategy) {
   return SUCCESS;
 }
 
+Status DropoutInfo::CheckStrategy(const StrategyPtr &strategy) {
+  if (CheckStrategyValue(strategy, inputs_shape_, is_auto_parallel_) != SUCCESS) {
+    if (is_auto_parallel_) {
+      MS_LOG(DEBUG) << name_ << " : Invalid strategy.";
+    } else {
+      MS_LOG(ERROR) << name_ << " : Invalid strategy.";
+    }
+    return FAILED;
+  }
+
+  // dropout don't support repeated calculation
+  CheckGlobalDeviceManager();
+  auto input_strategy = strategy->GetInputDim().at(0);
+  size_t dev_num = g_device_manager->GetDeviceListByStageId(0).size();
+  auto product_p = std::accumulate(input_strategy.begin(), input_strategy.end(), 1, std::multiplies<int>());
+  if (IntToSize(product_p) != dev_num) {
+    MS_LOG(ERROR) << name_ << ": Invalid strategy. Don't support repeated calc.";
+    return FAILED;
+  }
+
+  return SUCCESS;
+}
+
 Status ActivationInfo::GetAttrs() {
   if (attrs_.size() < ACTIVATION_ATTR_SIZE) {
     MS_LOG(ERROR) << name_ << " : The size of attrs small than 1.";
@@ -120,6 +145,27 @@ Status Activation::GenerateStrategies(int32_t stage_id) {
   return SUCCESS;
 }
 
+Status DropoutInfo::GenerateStrategies(int32_t stage_id) {
+  is_auto_parallel_ = true;
+  Shape input0_split(inputs_shape_[0].size(), 1);
+  Shapes splittable_inputs = {input0_split};
+
+  std::vector<StrategyPtr> sp_vector;
+  if (GenerateStrategiesForIndependentInputs(stage_id, inputs_shape_, splittable_inputs, &sp_vector) != SUCCESS) {
+    MS_LOG(ERROR) << name_ << " : Generate strategies for independent inputs() failed.";
+    return FAILED;
+  }
+  size_t success = 0;
+  for (auto &sp : sp_vector) {
+    if (SetCostUnderStrategy(sp) == SUCCESS) {
+      success++;
+      MS_LOG(INFO) << name_ << " : Successfully generated " << success << " strategy";
+      PrintStrategy(sp);
+    }
+  }
+  return SUCCESS;
+}
+
 Status Softmax::CheckStrategy(const StrategyPtr &strategy) {
   if (CheckStrategyValue(strategy, inputs_shape_, is_auto_parallel_) != SUCCESS) {
     if (is_auto_parallel_) {
@@ -334,6 +380,32 @@ Status ActivationBase::InferTensorInfo() {
   return SUCCESS;
 }
 
+Status DropoutInfo::InferTensorInfo() {
+  // infer tensor shape
+  Shape input_shape = inputs_shape_.at(0);
+
+  // infer slice shape
+  Shapes inputs_slice_shape, outputs_slice_shape;
+  Strategys inputs_strategy = strategy_->GetInputDim();
+  // dropout has two outputs
+  Strategys outputs_strategy = {inputs_strategy.at(0), inputs_strategy.at(0)};
+  if (InferSliceShape(inputs_strategy, outputs_strategy, &inputs_slice_shape, &outputs_slice_shape) != SUCCESS) {
+    return FAILED;
+  }
+  Shape input_slice_shape = inputs_slice_shape.at(0);
+  TensorLayout input_tensor_layout;
+  if (input_tensor_layout.InitFromVector(dev_matrix_shape_, inputs_tensor_map_[0], input_shape) != SUCCESS) {
+    return FAILED;
+  }
+  TensorInfo input_tensor_info(input_tensor_layout, input_shape, input_slice_shape);
+  inputs_tensor_info_.push_back(input_tensor_info);
+  // the two outputs of dropout all have the same tensor_info as input
+  outputs_tensor_info_.push_back(input_tensor_info);
+  outputs_tensor_info_.push_back(input_tensor_info);
+
+  return SUCCESS;
+}
+
 Status ActivationBase::Init(const StrategyPtr &strategy) {
   if (InitWithAutoRepeatCalc(strategy) != SUCCESS) {
     MS_LOG(ERROR) << name_ << " : Init failed.";

mindspore/ccsrc/frontend/parallel/ops_info/activation_info.h

@@ -219,6 +219,20 @@ class SigmoidInfo : public ActivationOther {
       : ActivationOther(name, inputs_shape, outputs_shape, attrs) {}
   ~SigmoidInfo() override = default;
 };
+
+class DropoutInfo : public ActivationOther {
+ public:
+  DropoutInfo(const std::string &name, const Shapes &inputs_shape, const Shapes &outputs_shape,
+              const PrimitiveAttrs &attrs)
+      : ActivationOther(name, inputs_shape, outputs_shape, attrs) {}
+  ~DropoutInfo() override = default;
+  Status GenerateStrategies(int32_t stage_id) override;
+
+ protected:
+  Status CheckStrategy(const StrategyPtr &strategy) override;
+  Status GetAttrs() override { return SUCCESS; }
+  Status InferTensorInfo() override;
+};
 }  // namespace parallel
 }  // namespace mindspore
 #endif  // MINDSPORE_CCSRC_FRONTEND_PARALLEL_OPS_INFO_ACTIVATION_INFO_H_

mindspore/ccsrc/frontend/parallel/ops_info/ops_utils.h

@@ -238,6 +238,7 @@ constexpr char UNSORTEF_SEGMENT_PRODD[] = "UnsortedSegmentProdD";
 constexpr char DEPTHWISE_CONV2D_NATIVE[] = "DepthwiseConv2dNative";
 constexpr char DEPTHWISE_CONV2D[] = "DepthwiseConv2D";
 constexpr char ADD[] = "Add";
+constexpr char DROPOUT[] = "Dropout";
 constexpr char KStridedSlice[] = "StridedSlice";
 
 // Parallel don't care

mindspore/ccsrc/frontend/parallel/step_auto_parallel.cc

@@ -256,7 +256,7 @@ bool IsSplittableOperator(const std::string &op_name) {
      REDUCE_MAX, REDUCE_MIN, ARGMAXWITHVALUE, ARGMINWITHVALUE, REDUCE_SUM, CONV2D, FUSE_BATCH_NORM, POOLING,
      MAX_POOL_WITH_ARGMAX, SIMPLE_MEAN, FLATTEN, BATCH_NORM, LAYER_NORM, BIAS_ADD, ASSIGN_SUB, COS, ACOS, EXP,
      LOG, REDUCE_MEAN, REAL_DIV, SIGMOID, POW, MAXIMUM, MINIMUM, EQUAL, NOT_EQUAL, LOGICALNOT, GATHERV2, SQRT,
-     STRIDEDSLICE, GET_NEXT, CAST, NEG, SQUARE, BATCH_MATMUL, EXPAND_DIMS, SQUEEZE, SPARSE_GATHERV2, TILE,
+     STRIDEDSLICE, GET_NEXT, CAST, NEG, SQUARE, BATCH_MATMUL, EXPAND_DIMS, SQUEEZE, SPARSE_GATHERV2, TILE, DROPOUT,
      SOFTMAX_CROSS_ENTROPY_WITH_LOGITS, SIGMOID_CROSS_ENTROPY_WITH_LOGITS, SPARSE_SOFTMAX_CROSS_ENTROPY_WITH_LOGITS};
   // clang-format on
 

tests/ut/python/parallel/test_gpu_dropout.py

@@ -0,0 +1,99 @@
+# 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 numpy as np
+import mindspore as ms
+import mindspore.nn as nn
+from mindspore import Tensor
+from mindspore import context
+from mindspore.common.api import _executor
+from mindspore.ops import composite as C
+from mindspore.ops import operations as P
+from tests.ut.python.ops.test_math_ops import VirtualLoss
+
+
+class NetWithLoss(nn.Cell):
+    def __init__(self, network):
+        super(NetWithLoss, self).__init__()
+        self.loss = VirtualLoss()
+        self.network = network
+
+    def construct(self, x, y):
+        predict = self.network(x, y)
+        return self.loss(predict)
+
+
+class GradWrap(nn.Cell):
+    def __init__(self, network):
+        super(GradWrap, self).__init__()
+        self.network = network
+
+    def construct(self, x, y):
+        return C.grad_all(self.network)(x, y)
+
+
+class Net(nn.Cell):
+    def __init__(self, strategy1=None, strategy2=None):
+        super().__init__()
+        self.dropout = P.Dropout(keep_prob=0.6).set_strategy(strategy1)
+        self.matmul = P.MatMul().set_strategy(strategy2)
+
+    def construct(self, x, y):
+        out = self.matmul(x, y)
+        out, _ = self.dropout(out)
+        return out
+
+
+def test_dropout_semi_auto():
+    context.set_auto_parallel_context(device_num=8, global_rank=0, parallel_mode="semi_auto_parallel")
+    net = GradWrap(NetWithLoss(Net()))
+    net.set_auto_parallel()
+
+    x = Tensor(np.ones([64, 32]), dtype=ms.float32)
+    y = Tensor(np.ones([32, 128]), dtype=ms.float32)
+    _executor.compile(net, x, y)
+
+
+def test_dropout_semi_auto2():
+    context.set_auto_parallel_context(device_num=8, global_rank=0, parallel_mode="semi_auto_parallel")
+    strategy1 = ((8, 1),)
+    strategy2 = ((4, 2), (2, 1))
+    net = GradWrap(NetWithLoss(Net(strategy1, strategy2)))
+    net.set_auto_parallel()
+
+    x = Tensor(np.ones([64, 32]), dtype=ms.float32)
+    y = Tensor(np.ones([32, 128]), dtype=ms.float32)
+    _executor.compile(net, x, y)
+
+
+def test_dropout_semi_auto3():
+    context.set_auto_parallel_context(device_num=8, global_rank=0, parallel_mode="semi_auto_parallel")
+    strategy1 = ((2, 4),)
+    strategy2 = ((4, 2), (2, 1))
+    net = GradWrap(NetWithLoss(Net(strategy1, strategy2)))
+    net.set_auto_parallel()
+
+    x = Tensor(np.ones([64, 32]), dtype=ms.float32)
+    y = Tensor(np.ones([32, 128]), dtype=ms.float32)
+    _executor.compile(net, x, y)
+
+
+def test_dropout_auto():
+    context.set_auto_parallel_context(device_num=8, global_rank=0, parallel_mode="auto_parallel")
+    net = GradWrap(NetWithLoss(Net()))
+    net.set_auto_parallel()
+
+    x = Tensor(np.ones([64, 32]), dtype=ms.float32)
+    y = Tensor(np.ones([32, 128]), dtype=ms.float32)
+    _executor.compile(net, x, y)