Support the 'drop_empty_grad' in of output of backward_ops (#49588)

* support the drop_empty_grad in backward

* change code according to yunfei's review suggestion

paddle/fluid/operators/generator/generate_op.py

@@ -459,6 +459,27 @@ def process_invoke_op(forward_op_dict, backward_op_dict):
                     )
 
 
+def parse_drop_empty_grad(op_fluid_list: list, bw_op_dict: dict):
+    for op_op in op_fluid_list:
+        if 'drop_empty_grad' in op_op:
+            bw_names = [
+                bw_name.split('(')[0].strip()
+                for bw_name in op_op['backward'].split(',')
+            ]
+            for bw_name in bw_names:
+                assert (
+                    bw_name in bw_op_dict
+                ), f"backward {bw_name} is not existed"
+                for out_grad in op_op['drop_empty_grad']:
+                    assert (
+                        out_grad in bw_op_dict[bw_name]['output_dict']
+                    ), f'''
+                         {bw_name} with {out_grad} is not existed in output_dict '''
+                    bw_op_dict[bw_name]['output_dict'][out_grad][
+                        'drop_empty_grad'
+                    ] = False
+
+
 def main(
     ops_yaml_path,
     backward_yaml_path,
@@ -488,6 +509,11 @@ def main(
         op['op_name'] = op['name']
     for bw_op in backward_ops:
         bw_op['op_name'] = bw_op['name']
+        for bw_output in bw_op['outputs']:
+            bw_output['drop_empty_grad'] = True
+
+    # deal the drop_empty_grad of bw_op by op_compat.yaml
+    parse_drop_empty_grad(op_fluid_map_list, backward_op_dict)
 
     parse_composite_info(ops, backward_ops, backward_op_dict)
 

paddle/fluid/operators/generator/templates/operator_utils.c.j2

@@ -417,7 +417,8 @@ class {{name | to_pascal_case}}OpMaker : public framework::SingleGradOpMaker<T>
       forward_input_names,
       forward_output_names,
       forward_input_orig_names,
-      forward_output_orig_names)}});
+      forward_output_orig_names,
+      output['drop_empty_grad'])}});
   {% endfor %}
 
     grad_op->SetAttrMap(this->Attrs());
@@ -480,7 +481,8 @@ class {{name | to_pascal_case}}OpMaker : public framework::SingleGradOpMaker<T>
       forward_input_names,
       forward_output_names,
       forward_input_orig_names,
-      forward_output_orig_names)}});
+      forward_output_orig_names,
+      true)}});
   {% endfor %}
 
   {% for attr in invoke_op["attrs"] %}
@@ -659,11 +661,15 @@ OutputGrad({{name_in_forward_orig | to_opmaker_name}})
 {%- endmacro %}
 
 {% macro extract_output_from_forward(name, input_names, output_names,
-  input_orig_names, output_orig_names) %}{# inline #}
+  input_orig_names, output_orig_names, drop_empty_grad) %}{# inline #}
   {% if name[:-5] in input_names %}
     {% set name_in_forward = name[:-5] %}
     {% set name_in_forward_orig = input_orig_names[input_names.index(name_in_forward)]%}
+      {%- if drop_empty_grad is true -%}
 InputGrad({{name_in_forward_orig | to_opmaker_name}})
+      {%- else -%}
+InputGrad({{name_in_forward_orig | to_opmaker_name}}, false)
+      {%- endif %}
   {%- elif (name) in input_names %}
     {% set name_in_forward_orig = input_orig_names[input_names.index(name)]%}
 Input({{name  | to_opmaker_name}})

paddle/fluid/operators/meshgrid_op.cc

-// Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
-//
-// 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 <memory>
-#include <string>
-#include <vector>
-
-#include "paddle/fluid/framework/infershape_utils.h"
-#include "paddle/fluid/framework/op_registry.h"
-#include "paddle/fluid/framework/operator.h"
-#include "paddle/phi/core/infermeta_utils.h"
-#include "paddle/phi/infermeta/multiary.h"
-
-namespace paddle {
-namespace operators {
-
-class MeshgridOp : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-
- protected:
-  phi::KernelKey GetExpectedKernelType(
-      const framework::ExecutionContext& ctx) const override {
-    auto inputs = ctx.MultiInput<phi::DenseTensor>("X");
-    auto input_data_type = framework::proto::VarType::Type(0);
-    bool flag = 0;
-    for (auto* input : inputs) {
-      if (input->IsInitialized() && input->numel() > 0) {
-        input_data_type = framework::TransToProtoVarType(input->dtype());
-        flag = 1;
-        break;
-      }
-    }
-    if (flag == 0) {
-      PADDLE_THROW(platform::errors::InvalidArgument(
-          "All Inputs of Meshgrid OP are Empty!"));
-    }
-
-    return phi::KernelKey(input_data_type, ctx.GetPlace());
-  }
-};
-
-class MeshgridOpMaker : public framework::OpProtoAndCheckerMaker {
- public:
-  void Make() override {
-    AddInput("X", "(Tensor, default Tensor<float>).").AsDuplicable();
-    AddOutput("Out", "(Tensor, default Tensor<float>.)").AsDuplicable();
-
-    AddComment(R"DOC(
-Meshgrid Operator.
-Take: N tensors, each of which can be either scalr or 1-dimensional vector, and create
-N-dimensional grids.
-
-Args:
-  tensors (list of tensor): if the input k tensors has (N1,), (N2,),..., (Nk,), then
-  the output tensors are all of size (N1, N2, ...., Nk).
-
-Example::
->>> x = fluid.data(name='x', shape=[10], dtype='float64')
->>> y = fluid.data(name='y', shape=[20], dtype='float64')
->>> grid_x, grid_y = fluid.layers.meshgrid([x, y])
->>> grid_x.shape
-(10,20)
->>> grid_y.shape
-(10,20)
-)DOC");
-  }
-};
-
-class MeshgridGradOp : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-
- protected:
-  void InferShape(framework::InferShapeContext* ctx) const override {
-    PADDLE_ENFORCE_GT(ctx->Inputs(framework::GradVarName("Out")).size(),
-                      1,
-                      platform::errors::InvalidArgument(
-                          "Number of Inputs(Out@Grad) should be larger than 1."
-                          "But received Inputs(Out@Grad)' size = %d .",
-                          ctx->Inputs(framework::GradVarName("Out")).size()));
-    ctx->SetOutputsDim(framework::GradVarName("X"), ctx->GetInputsDim("X"));
-  }
-
- protected:
-  phi::KernelKey GetExpectedKernelType(
-      const framework::ExecutionContext& ctx) const override {
-    return phi::KernelKey(OperatorWithKernel::IndicateVarDataType(
-                              ctx, framework::GradVarName("Out")),
-                          ctx.device_context().GetPlace());
-  }
-};
-
-template <typename T>
-class MeshgridGradOpMaker : public framework::SingleGradOpMaker<T> {
- public:
-  using framework::SingleGradOpMaker<T>::SingleGradOpMaker;
-
- protected:
-  void Apply(GradOpPtr<T> op) const override {
-    op->SetType("meshgrid_grad");
-    op->SetInput("X", this->Input("X"));
-    op->SetInput(framework::GradVarName("Out"), this->OutputGrad("Out"));
-    op->SetOutput(framework::GradVarName("X"), this->InputGrad("X", false));
-  }
-};
-
-}  // namespace operators
-}  // namespace paddle
-
-namespace ops = paddle::operators;
-DECLARE_INFER_SHAPE_FUNCTOR(meshgrid,
-                            MeshgridInferShapeFunctor,
-                            PD_INFER_META(phi::MeshgridInferMeta));
-REGISTER_OPERATOR(meshgrid,
-                  ops::MeshgridOp,
-                  ops::MeshgridOpMaker,
-                  ops::MeshgridGradOpMaker<paddle::framework::OpDesc>,
-                  ops::MeshgridGradOpMaker<paddle::imperative::OpBase>,
-                  MeshgridInferShapeFunctor);
-REGISTER_OPERATOR(meshgrid_grad, ops::MeshgridGradOp);

paddle/fluid/operators/multi_dot_op.cc

-/* Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
-
-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 <algorithm>
-#include <utility>
-#include <vector>
-
-#include "paddle/fluid/framework/infershape_utils.h"
-#include "paddle/fluid/framework/op_registry.h"
-#include "paddle/fluid/framework/op_version_registry.h"
-#include "paddle/fluid/operators/strided_memcpy.h"
-#include "paddle/fluid/operators/utils.h"
-#include "paddle/phi/core/infermeta_utils.h"
-#include "paddle/phi/infermeta/multiary.h"
-#include "paddle/phi/kernels/funcs/blas/blas.h"
-
-namespace paddle {
-namespace operators {
-
-class MultiDotOpMaker : public framework::OpProtoAndCheckerMaker {
- public:
-  void Make() override {
-    AddInput("X", "The input tensors of multi_dot operator.").AsDuplicable();
-    AddOutput("Out", "The output tensor of multi_dot operator");
-    AddComment(R"DOC(
-Compute the dot product of two or more arrays in a single function call, while automatically selecting the fastest evaluation order.
-
-multi_dot chains MatMul and uses optimal parenthesization of the matrices [1] [2]. Depending on the shapes of the matrices, this can speed up the multiplication a lot.
-
-If the first argument is 1-D it is treated as a row vector. If the last argument is 1-D it is treated as a column vector. The other arguments must be 2-D.
-      )DOC");
-  }
-};
-
-class MultiDotOp : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-};
-
-class MultiDotOpGrad : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-
-  void InferShape(framework::InferShapeContext* ctx) const override {
-    OP_INOUT_CHECK(ctx->HasInputs("X"), "Input", "X", "multi_dot");
-    OP_INOUT_CHECK(ctx->HasInput(framework::GradVarName("Out")),
-                   "Input",
-                   "Out@GRAD",
-                   "multi_dot");
-
-    auto in_x = "X";
-    auto out_x_g_n = framework::GradVarName(in_x);
-    auto ins_dims = ctx->GetInputsDim(in_x);
-    ctx->SetOutputsDim(out_x_g_n, ins_dims);
-    ctx->ShareAllLoD(in_x, out_x_g_n);
-  }
-};
-
-template <typename T>
-class MultiDotOpGradMaker : public framework::SingleGradOpMaker<T> {
- public:
-  using framework::SingleGradOpMaker<T>::SingleGradOpMaker;
-
- protected:
-  void Apply(GradOpPtr<T> op) const override {
-    op->SetType("multi_dot_grad");
-    op->SetInput("X", this->Input("X"));
-    op->SetInput(framework::GradVarName("Out"), this->OutputGrad("Out"));
-    op->SetOutput(framework::GradVarName("X"), this->InputGrad("X", false));
-  }
-};
-template <typename T>
-class MultiDotOpDoubleGradMaker : public framework::SingleGradOpMaker<T> {
- public:
-  using framework::SingleGradOpMaker<T>::SingleGradOpMaker;
-
- protected:
-  void Apply(GradOpPtr<T> grad_op) const override {
-    grad_op->SetType("multi_dot");
-    grad_op->SetInput("X", this->Input(("X")));
-    grad_op->SetInput("DOut", this->Input(framework::GradVarName("Out")));
-    grad_op->SetOutput("DDx", this->OutputGrad(framework::GradVarName("X")));
-  }
-};
-
-}  // namespace operators
-}  // namespace paddle
-
-namespace ops = paddle::operators;
-
-DECLARE_INFER_SHAPE_FUNCTOR(multi_dot,
-                            MultiDotInferShapeFunctor,
-                            PD_INFER_META(phi::MultiDotInferMeta));
-
-REGISTER_OPERATOR(multi_dot,
-                  ops::MultiDotOp,
-                  ops::MultiDotOpMaker,
-                  ops::MultiDotOpGradMaker<paddle::framework::OpDesc>,
-                  ops::MultiDotOpGradMaker<paddle::imperative::OpBase>,
-                  MultiDotInferShapeFunctor);
-
-REGISTER_OPERATOR(multi_dot_grad,
-                  ops::MultiDotOpGrad,
-                  ops::MultiDotOpDoubleGradMaker<paddle::framework::OpDesc>,
-                  ops::MultiDotOpDoubleGradMaker<paddle::imperative::OpBase>);

paddle/fluid/operators/multiplex_op.cc

-/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
-
-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 <memory>
-#include <vector>
-
-#include "paddle/fluid/framework/infershape_utils.h"
-#include "paddle/fluid/framework/op_registry.h"
-#include "paddle/phi/core/infermeta_utils.h"
-#include "paddle/phi/infermeta/multiary.h"
-
-namespace paddle {
-namespace operators {
-
-class MultiplexOp : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-
- protected:
-  phi::KernelKey GetExpectedKernelType(
-      const framework::ExecutionContext& ctx) const override {
-    return phi::KernelKey(OperatorWithKernel::IndicateVarDataType(ctx, "X"),
-                          ctx.device_context().GetPlace());
-  }
-};
-
-class MultiplexOpMaker : public framework::OpProtoAndCheckerMaker {
- public:
-  void Make() override {
-    AddInput("Ids",
-             "Tensor<int32>, index variable which is a 2-D tensor with shape "
-             "[M, 1] where M is the batch size.");
-    AddInput("X",
-             "A list of variables to gather from. All variables have the same "
-             "shape and the rank is at least 2.")
-        .AsDuplicable();
-    AddOutput("Out", "The output tensor of multiplex operator.");
-    AddComment(R"DOC(
-Referring to the given index variable, this layer selects rows from the
-input variables to construct a multiplex variable. Assuming that there are
-:math:`m` input variables and :math:`I_i` represents the i-th input
-variable and :math:`i` is in [0, :math:`m`). All input variables are
-tensors with same shape [:math:`d_0`, :math:`d_1`, ..., :math:`d_R`].
-Please note that rank of the input tensor should be at least 2. Each input
-variable will be treated as a 2-D matrix with shape [:math:`M`, :math:`N`]
-where :math:`M` for :math:`d_0` and :math:`N` for :math:`d_1` * :math:`d_2`
-* ... * :math:`d_R`. Let :math:`I_i[j]` be the j-th row of the i-th input
-variable. The given index variable should be a 2-D tensor with shape
-[:math:`M`, 1]. Let `ID[i]` be the i-th index value of the index variable.
-Then the output variable will be a tensor with shape [:math:`d_0`,
-:math:`d_1`, ..., :math:`d_R`]. If we treat the output tensor as a 2-D
-matrix with shape [:math:`M`, :math:`N`] and let :math:`O[i]` be the i-th
-row of the matrix, then `O[i]` is equal to :math:`I_{ID[i]}[i]`.
-
-* Ids: the index tensor.
-
-* X[0 : N - 1]: the candidate tensors for output (N >= 2).
-
-* For each index i from 0 to batchSize - 1, the output is the i-th row of the
-the (Ids[i])-th tensor.
-
-For i-th row of the output tensor:
-
-$$
-y[i] = x_{k}[i]
-$$
-
-where $y$ is the output tensor, $x_{k}$ is the k-th input tensor,
-and $k = Ids[i]$.
-
-)DOC");
-  }
-};
-
-class MultiplexGradOp : public framework::OperatorWithKernel {
- public:
-  using framework::OperatorWithKernel::OperatorWithKernel;
-
-  void InferShape(framework::InferShapeContext* ctx) const override {
-    auto dxs = ctx->Outputs(framework::GradVarName("X"));
-    PADDLE_ENFORCE_NE(dxs.empty(),
-                      true,
-                      platform::errors::InvalidArgument(
-                          "Output(X@Grad) should not be null."));
-    OP_INOUT_CHECK(ctx->HasInput(framework::GradVarName("Out")),
-                   "Input",
-                   framework::GradVarName("Out"),
-                   "MultiplexGrad");
-    auto dout_dim = ctx->GetInputDim(framework::GradVarName("Out"));
-    ctx->SetOutputsDim(framework::GradVarName("X"),
-                       std::vector<framework::DDim>(dxs.size(), dout_dim));
-  }
-
- protected:
-  phi::KernelKey GetExpectedKernelType(
-      const framework::ExecutionContext& ctx) const override {
-    return phi::KernelKey(OperatorWithKernel::IndicateVarDataType(
-                              ctx, framework::GradVarName("Out")),
-                          ctx.device_context().GetPlace());
-  }
-};
-
-template <typename T>
-class MultiplexGradMaker : public framework::SingleGradOpMaker<T> {
- public:
-  using framework::SingleGradOpMaker<T>::SingleGradOpMaker;
-
- protected:
-  void Apply(GradOpPtr<T> op) const override {
-    op->SetType("multiplex_grad");
-    op->SetInput("Ids", this->Input("Ids"));
-    op->SetInput(framework::GradVarName("Out"), this->OutputGrad("Out"));
-    op->SetOutput(framework::GradVarName("X"), this->InputGrad("X", false));
-    op->SetAttrMap(this->Attrs());
-  }
-};
-
-}  // namespace operators
-}  // namespace paddle
-
-namespace ops = paddle::operators;
-DECLARE_INFER_SHAPE_FUNCTOR(multiplex,
-                            MultiplexInferShapeFunctor,
-                            PD_INFER_META(phi::MultiplexInferMeta));
-
-REGISTER_OPERATOR(multiplex,
-                  ops::MultiplexOp,
-                  ops::MultiplexOpMaker,
-                  ops::MultiplexGradMaker<paddle::framework::OpDesc>,
-                  ops::MultiplexGradMaker<paddle::imperative::OpBase>,
-                  MultiplexInferShapeFunctor);
-REGISTER_OPERATOR(multiplex_grad, ops::MultiplexGradOp);

paddle/phi/api/yaml/backward.yaml

@@ -841,6 +841,16 @@
   kernel :
     func : maxout_grad
 
+- backward_op : meshgrid_grad
+  forward : meshgrid (Tensor[] inputs) -> Tensor[](outputs)
+  args : (Tensor[] inputs, Tensor[] outputs_grad)
+  output : Tensor[](inputs_grad){inputs.size()}
+  infer_meta :
+    func : MeshgridGradInferMeta
+  kernel :
+    func : meshgrid_grad
+    data_type : outputs_grad
+
 - backward_op : mode_grad
   forward : mode(Tensor x,  int axis = -1,  bool keepdim = false) -> Tensor(out), Tensor(indices)
   args : (Tensor x, Tensor indices, Tensor out_grad,  int axis,  bool keepdim)
@@ -851,6 +861,27 @@
   kernel :
     func : mode_grad
 
+- backward_op : multi_dot_grad
+  forward : multi_dot (Tensor[] x) -> Tensor(out)
+  args : (Tensor[] x, Tensor out_grad)
+  output : Tensor[](x_grad) {x.size()}
+  infer_meta :
+    func : MultiDotGradInferMeta
+  kernel :
+    func : multi_dot_grad
+
+- backward_op : multiplex_grad
+  forward : multiplex (Tensor[] inputs, Tensor index) -> Tensor(out)
+  args : (Tensor[] inputs, Tensor index, Tensor out_grad)
+  output : Tensor[](inputs_grad){inputs.size()}
+  infer_meta :
+    func : MultiplexGradInferMeta
+    param : [index, out_grad]
+  kernel :
+    func : multiplex_grad
+    param : [index, out_grad]
+    data_type : out_grad
+
 - backward_op : mv_grad
   forward : mv (Tensor x, Tensor vec) -> Tensor(out)
   args : (Tensor x, Tensor vec, Tensor out_grad)

paddle/phi/api/yaml/legacy_backward.yaml

@@ -806,15 +806,6 @@
   backward : mean_double_grad
   no_need_buffer : x
 
-- backward_op : meshgrid_grad
-  forward : meshgrid (Tensor[] inputs) -> Tensor[](outputs)
-  args : (Tensor[] inputs, Tensor[] outputs_grad)
-  output : Tensor[](inputs_grad){inputs.size()}
-  infer_meta :
-    func : MeshgridGradInferMeta
-  kernel :
-    func : meshgrid_grad
-
 - backward_op : min_grad
   forward: min (Tensor x,  IntArray axis={},  bool keepdim=false) -> Tensor(out)
   args : (Tensor x, Tensor out, Tensor out_grad, IntArray axis={}, bool keepdim=false, bool reduce_all=false)
@@ -846,26 +837,6 @@
     func : mish_grad
   inplace : (out_grad -> x_grad)
 
-- backward_op : multi_dot_grad
-  forward : multi_dot (Tensor[] x) -> Tensor(out)
-  args : (Tensor[] x, Tensor out_grad)
-  output : Tensor[](x_grad) {x.size()}
-  infer_meta :
-    func : MultiDotGradInferMeta
-  kernel :
-    func : multi_dot_grad
-
-- backward_op : multiplex_grad
-  forward : multiplex (Tensor[] inputs, Tensor index) -> Tensor(out)
-  args : (Tensor[] inputs, Tensor index, Tensor out_grad)
-  output : Tensor[](inputs_grad){inputs.size()}
-  infer_meta :
-    func : MultiplexGradInferMeta
-    param : [index, out_grad]
-  kernel :
-    func : multiplex_grad
-    param : [index, out_grad]
-
 - backward_op : multiply_double_grad
   forward : multiply_grad (Tensor x, Tensor y, Tensor grad_out, int axis = -1) -> Tensor(grad_x), Tensor(grad_y)
   args : (Tensor x, Tensor y, Tensor grad_out, Tensor grad_x_grad, Tensor grad_y_grad, int axis = -1)

paddle/phi/api/yaml/legacy_ops.yaml

@@ -1177,15 +1177,6 @@
     data_type : param
   inplace : (param -> param_out), (velocity -> velocity_out), (master_param -> master_param_out)
 
-- op : meshgrid
-  args : (Tensor[] inputs)
-  output : Tensor[]{inputs.size()}
-  infer_meta :
-    func : MeshgridInferMeta
-  kernel :
-    func : meshgrid
-  backward : meshgrid_grad
-
 - op : min
   args : (Tensor x, IntArray axis={}, bool keepdim=false)
   output : Tensor(out)
@@ -1225,15 +1216,6 @@
   optional : master_param
   inplace : (param -> param_out), (velocity -> velocity_out), (master_param -> master_param_out)
 
-- op : multi_dot
-  args : (Tensor[] x)
-  output : Tensor
-  infer_meta :
-    func : MultiDotInferMeta
-  kernel :
-    func : multi_dot
-  backward : multi_dot_grad
-
 - op : multiclass_nms3
   args : (Tensor bboxes, Tensor scores, Tensor rois_num, float score_threshold, int nms_top_k, int keep_top_k, float nms_threshold=0.3, bool normalized=true, float nms_eta=1.0, int background_label=0)
   output : Tensor(out), Tensor(index), Tensor(nms_rois_num)
@@ -1243,16 +1225,6 @@
     func : multiclass_nms3
   optional : rois_num
 
-- op : multiplex
-  args : (Tensor[] inputs, Tensor index)
-  output : Tensor
-  infer_meta :
-    func : MultiplexInferMeta
-  kernel :
-    func : multiplex
-    data_type : inputs
-  backward : multiplex_grad
-
 - op : multiply
   args : (Tensor x, Tensor y)
   output : Tensor

paddle/phi/api/yaml/op_compat.yaml

@@ -919,6 +919,14 @@
   outputs :
     out : Out
 
+- op : meshgrid
+  backward : meshgrid_grad
+  inputs :
+    inputs : X
+  outputs :
+    out : Out
+  drop_empty_grad : [inputs_grad]
+
 - op : mish
   backward : mish_grad
   extra :
@@ -931,6 +939,14 @@
   outputs :
     {out : Out, indices : Indices}
 
+- op : multi_dot
+  backward : multi_dot_grad
+  inputs :
+    x : X
+  outputs :
+    out : Out
+  drop_empty_grad : [x_grad]
+
 - op : multinomial
   inputs :
     {x : X}
@@ -941,6 +957,14 @@
       data_type : int
       support_tensor : true
 
+- op : multiplex
+  backward : multiplex_grad
+  inputs :
+    {inputs : X, index : Ids}
+  outputs :
+    out : Out
+  drop_empty_grad : [inputs_grad]
+
 - op : multiply (elementwise_mul)
   backward : multiply_grad (elementwise_mul_grad)
   extra :

paddle/phi/api/yaml/ops.yaml

@@ -795,6 +795,16 @@
     func : maxout
   backward : maxout_grad
 
+- op : meshgrid
+  args : (Tensor[] inputs)
+  output : Tensor[]{inputs.size()}
+  infer_meta :
+    func : MeshgridInferMeta
+  kernel :
+    func : meshgrid
+    data_type : inputs
+  backward : meshgrid_grad
+
 - op : mode
   args : (Tensor x,  int axis = -1,  bool keepdim = false)
   output : Tensor(out), Tensor(indices)
@@ -804,6 +814,15 @@
     func : mode
   backward : mode_grad
 
+- op : multi_dot
+  args : (Tensor[] x)
+  output : Tensor
+  infer_meta :
+    func : MultiDotInferMeta
+  kernel :
+    func : multi_dot
+  backward : multi_dot_grad
+
 - op : multinomial
   args : (Tensor x, Scalar(int) num_samples = 1, bool replacement = false)
   output : Tensor(out)
@@ -813,6 +832,16 @@
     func : multinomial
     data_type : x
 
+- op : multiplex
+  args : (Tensor[] inputs, Tensor index)
+  output : Tensor
+  infer_meta :
+    func : MultiplexInferMeta
+  kernel :
+    func : multiplex
+    data_type : inputs
+  backward : multiplex_grad
+
 - op : mv
   args : (Tensor x, Tensor vec)
   output : Tensor

paddle/phi/infermeta/backward.cc

@@ -658,7 +658,9 @@ void MultiplexGradInferMeta(const MetaTensor& ids,
       errors::InvalidArgument("Output(X@Grad) should not be null."));
   auto dout_dim = out_grad.dims();
   for (auto in_grad : ins_grad) {
-    in_grad->set_dims(dout_dim);
+    if (in_grad != nullptr) {
+      in_grad->set_dims(dout_dim);
+    }
   }
 }
 

paddle/phi/ops/compat/meshgrid_sig.cc

-// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
-//
-// 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 "paddle/phi/core/compat/op_utils.h"
-
-namespace phi {
-
-KernelSignature MeshgridOpArgumentMapping(const ArgumentMappingContext& ctx) {
-  return KernelSignature("meshgrid", {"X"}, {}, {"Out"});
-}
-
-KernelSignature MeshgridGradOpArgumentMapping(
-    const ArgumentMappingContext& ctx) {
-  return KernelSignature("meshgrid_grad", {"X", "Out@GRAD"}, {}, {"X@GRAD"});
-}
-
-}  // namespace phi
-
-PD_REGISTER_ARG_MAPPING_FN(meshgrid, phi::MeshgridOpArgumentMapping);
-PD_REGISTER_ARG_MAPPING_FN(meshgrid_grad, phi::MeshgridGradOpArgumentMapping);

paddle/phi/ops/compat/multi_dot_sig.cc

-// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
-//
-// 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 "paddle/phi/core/compat/op_utils.h"
-
-namespace phi {
-
-KernelSignature MultiDotGradOpArgumentMapping(
-    const ArgumentMappingContext& ctx) {
-  return KernelSignature("multi_dot_grad", {"X", "Out@GRAD"}, {}, {"X@GRAD"});
-}
-
-}  // namespace phi
-
-PD_REGISTER_ARG_MAPPING_FN(multi_dot_grad, phi::MultiDotGradOpArgumentMapping);

paddle/phi/ops/compat/multiplex_sig.cc

-/* Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
-
-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 "paddle/phi/core/compat/op_utils.h"
-
-namespace phi {
-
-KernelSignature MultiplexOpArgumentMapping(const ArgumentMappingContext& ctx) {
-  return KernelSignature("multiplex", {"X", "Ids"}, {}, {"Out"});
-}
-
-KernelSignature MultiplexGradOpArgumentMapping(
-    const ArgumentMappingContext& ctx) {
-  return KernelSignature("multiplex_grad", {"Ids", "Out@GRAD"}, {}, {"X@GRAD"});
-}
-
-}  // namespace phi
-
-PD_REGISTER_ARG_MAPPING_FN(multiplex, phi::MultiplexOpArgumentMapping);
-PD_REGISTER_ARG_MAPPING_FN(multiplex_grad, phi::MultiplexGradOpArgumentMapping);