match_matrix_tensor_op.cc

/* Copyright (c) 2018 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 <fstream>
#include <iomanip>
#include <iostream>
#include <memory>
#include <vector>

#include "paddle/fluid/operators/match_matrix_tensor_op.h"
#include "paddle/fluid/operators/search_compute.h"

namespace paddle {
namespace operators {
using Tensor = framework::Tensor;
using LoDTensor = framework::LoDTensor;
using LoD = framework::LoD;

void MatchMatrixTensorOP::InferShape(framework::InferShapeContext* ctx) const {
  OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X", "match_matrix_tensor");
  OP_INOUT_CHECK(ctx->HasInput("Y"), "Input", "Y", "match_matrix_tensor");
  OP_INOUT_CHECK(ctx->HasInput("W"), "Input", "W", "match_matrix_tensor");
  OP_INOUT_CHECK(ctx->HasOutput("Out"), "Output", "Out", "match_matrix_tensor");
  OP_INOUT_CHECK(ctx->HasOutput("Tmp"), "Output", "Tmp", "match_matrix_tensor");

  auto x_dims = ctx->GetInputDim("X");
  PADDLE_ENFORCE_EQ(x_dims.size(), 2,
                    platform::errors::InvalidArgument(
                        "The dimensions of Input(X) should be equal to 2, "
                        "but received %d.",
                        x_dims.size()));

  auto y_dims = ctx->GetInputDim("Y");
  PADDLE_ENFORCE_EQ(y_dims.size(), 2,
                    platform::errors::InvalidArgument(
                        "The dimensions of Input(Y) should be equal to 2, "
                        "but received %d.",
                        y_dims.size()));

  auto w_dims = ctx->GetInputDim("W");
  PADDLE_ENFORCE_EQ(w_dims.size(), 3,
                    platform::errors::InvalidArgument(
                        "The dimensions of Input(W) should be equal to 3, "
                        "but received %d.",
                        w_dims.size()));

  int dim_t = ctx->Attrs().Get<int>("dim_t");
  PADDLE_ENFORCE_EQ(
      w_dims[0], x_dims[1],
      platform::errors::InvalidArgument(
          "The first dimension of Input(W) should be equal to the second "
          "dimension of Input(X). But received the first dimension of Input(W) "
          "is %d, the second dimension of Input(X) is %d.",
          w_dims[0], x_dims[1]));
  PADDLE_ENFORCE_EQ(
      w_dims[1], dim_t,
      platform::errors::InvalidArgument(
          "The second dimension of Input(W) should be equal to 'dim_t', but "
          "received the second dimension of Input(W) is %d, 'dim_t' is %d.",
          w_dims[1], dim_t));
  PADDLE_ENFORCE_EQ(
      w_dims[2], y_dims[1],
      platform::errors::InvalidArgument(
          "The last dimension of Input(W) should be equal to "
          "the second dimension of Input(Y). But received the last dimension "
          "of Input(W) is %d, the second dimension of Input(Y) is %d.",
          w_dims[2], y_dims[1]));

  int64_t out_dim_0 = -1;
  int64_t tmp_dim_0 = -1;
  if (ctx->IsRuntime()) {
    framework::Variable* x_var =
        BOOST_GET(framework::Variable*, ctx->GetInputVarPtrs("X")[0]);
    const auto& x_lod = x_var->Get<LoDTensor>().lod();
    PADDLE_ENFORCE_EQ(x_lod.empty(), false,
                      platform::errors::InvalidArgument(
                          "The Input(X) should hold LoD information, but "
                          "received Input(X).lod() is empty."));
    const auto& x_lod_0 = x_lod[0];
    PADDLE_ENFORCE_GE(x_lod_0.size(), 2,
                      platform::errors::InvalidArgument(
                          "The dimensions of Input(X)'s LoD data should be "
                          "equal to 2, but received %d.",
                          x_lod_0.size()));
    PADDLE_ENFORCE_EQ(x_dims[0], static_cast<int64_t>(x_lod_0.back()),
                      platform::errors::InvalidArgument(
                          "The last element of Input(X)'s LoD data should be "
                          "equal to the first dimension of Input(X). "
                          "But received the last element of Input(X)'s LoD "
                          "data is %d, the first dimension of Input(X) is %d.",
                          x_lod_0.back(), x_dims[0]));

    framework::Variable* y_var =
        BOOST_GET(framework::Variable*, ctx->GetInputVarPtrs("Y")[0]);
    const auto& y_lod = y_var->Get<LoDTensor>().lod();
    PADDLE_ENFORCE_EQ(y_lod.empty(), false,
                      platform::errors::InvalidArgument(
                          "The Input(Y) should hold LoD information, but "
                          "received Input(Y).lod() is empty."));
    const auto& y_lod_0 = y_lod[0];
    PADDLE_ENFORCE_GE(y_lod_0.size(), 2,
                      platform::errors::InvalidArgument(
                          "The dimensions of Input(Y)'s LoD data should be "
                          "equal to 2, but received %d.",
                          y_lod_0.size()));
    PADDLE_ENFORCE_EQ(y_dims[0], static_cast<int64_t>(y_lod_0.back()),
                      platform::errors::InvalidArgument(
                          "The last element of Input(Y)'s LoD data should be "
                          "equal to the first dimension of Input(Y). "
                          "But received the last element of Input(Y)'s LoD "
                          "data is %d, the first dimension of Input(Y) is %d.",
                          y_lod_0.back(), y_dims[0]));

    PADDLE_ENFORCE_EQ(x_lod_0.size(), y_lod_0.size(),
                      platform::errors::InvalidArgument(
                          "The dimensions of Input(X)'s and Input(Y)'s LoD "
                          "data should be equal. "
                          "But received the dimensions of Input(X)'s LoD is "
                          "%d, the dimensions of Input(Y)'s LoD is %d.",
                          x_lod_0.size(), y_lod_0.size()));

    out_dim_0 = 0;
    for (size_t i = 1; i < x_lod_0.size(); i++) {
      int64_t x_len = x_lod_0[i] - x_lod_0[i - 1];
      int64_t y_len = y_lod_0[i] - y_lod_0[i - 1];
      out_dim_0 += (x_len * y_len);
    }
    out_dim_0 *= dim_t;

    tmp_dim_0 = x_dims[0] * dim_t * x_dims[1];
  } else {
    // compile time
    framework::VarDesc* x_desc =
        BOOST_GET(framework::VarDesc*, ctx->GetInputVarPtrs("X")[0]);
    PADDLE_ENFORCE_GE(
        x_desc->GetLoDLevel(), 1,
        platform::errors::InvalidArgument("The LoD level of Input(X) should be "
                                          "greater than 1, but reviced %d.",
                                          x_desc->GetLoDLevel()));
    framework::VarDesc* y_desc =
        BOOST_GET(framework::VarDesc*, ctx->GetInputVarPtrs("Y")[0]);
    PADDLE_ENFORCE_GE(
        y_desc->GetLoDLevel(), 1,
        platform::errors::InvalidArgument("The LoD level of Input(Y) should be "
                                          "greater than 1, but reviced %d.",
                                          y_desc->GetLoDLevel()));
    ctx->ShareLoD("X", "Out");
  }

  std::vector<int64_t> out_dims_vec{out_dim_0};
  out_dims_vec.push_back(1);
  std::vector<int64_t> tmp_dims_vec{tmp_dim_0};
  tmp_dims_vec.push_back(1);
  ctx->SetOutputDim("Out", pten::make_ddim(out_dims_vec));
  ctx->SetOutputDim("Tmp", pten::make_ddim(tmp_dims_vec));
}

void MatchMatrixTensorOpGrad::InferShape(
    framework::InferShapeContext* ctx) const {
  OP_INOUT_CHECK(ctx->HasInput("X"), "Input", "X", "match_matrix_tensor_grad");
  OP_INOUT_CHECK(ctx->HasInput("Y"), "Input", "Y", "match_matrix_tensor_grad");
  OP_INOUT_CHECK(ctx->HasInput("W"), "Input", "W", "match_matrix_tensor_grad");
  OP_INOUT_CHECK(ctx->HasInput(framework::GradVarName("Out")), "Input",
                 "Out@GRAD", "match_matrix_tensor_grad");

  if (ctx->HasOutput(framework::GradVarName("X"))) {
    ctx->SetOutputDim(framework::GradVarName("X"), ctx->GetInputDim("X"));
    ctx->ShareLoD("X", /*->*/ framework::GradVarName("X"));
  }
  if (ctx->HasOutput(framework::GradVarName("Y"))) {
    ctx->SetOutputDim(framework::GradVarName("Y"), ctx->GetInputDim("Y"));
    ctx->ShareLoD("Y", /*->*/ framework::GradVarName("Y"));
  }
  if (ctx->HasOutput(framework::GradVarName("W"))) {
    ctx->SetOutputDim(framework::GradVarName("W"), ctx->GetInputDim("W"));
  }
}

void MatchMatrixTensorOpMaker::Make() {
  AddInput("X",
           "X (LoDTensor, default LoDTensor<float>) Input variable which "
           "should contain lod information.");
  AddInput("Y",
           "Y (LoDTensor, default LoDTensor<float>) Input variable which "
           "should contain lod information.");
  AddInput("W", "W (Tensor), The weight of X and Y.");
  AddAttr<int>("dim_t", "the dim of W").SetDefault(1);
  AddOutput("Out",
            "(LoDTensor, default LoDTensor<float>) Output variable which "
            "is X * W * Y");
  AddOutput("Tmp",
            "(LoDTensor, default LoDTensor<float>) tmp variable which is "
            "used for X * W");
  AddComment(R"DOC(
      Match Matrix Tensor Operator

      This operator calculate X * W * Y, only support 2-D for X and Y.
      the output is a level-1 LodTensor: 
        level_0: dim_t
      
      NOTE: only support 'float32' data type now.

    )DOC");
}

template <typename DeviceContext, typename T>
class CPUMatchMatrixTensorOPKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
    auto* x = ctx.Input<LoDTensor>("X");
    auto* y = ctx.Input<LoDTensor>("Y");
    auto* w = ctx.Input<Tensor>("W");
    auto* out = ctx.Output<LoDTensor>("Out");
    auto* tmp = ctx.Output<LoDTensor>("Tmp");

    int dim_t = ctx.Attr<int>("dim_t");
    int64_t dim_in = x->dims()[1];

    const auto& offset_l = x->lod()[0];
    const auto& offset_r = y->lod()[0];

    std::vector<size_t> top_offset;
    size_t top_size = 0;
    top_offset.push_back(top_size);
    for (size_t b = 0; b < x->lod()[0].size() - 1; b++) {
      size_t len_l = offset_l[b + 1] - offset_l[b];
      size_t len_r = offset_r[b + 1] - offset_r[b];
      top_size += dim_t * len_l * len_r;
      top_offset.push_back(top_size);
    }
    auto* out_data = out->mutable_data<T>(ctx.GetPlace());
    memset(out_data, 0.0, out->dims()[0] * out->dims()[1] * sizeof(T));

    auto* bottom_l_data = x->data<T>();
    auto* bottom_r_data = y->data<T>();
    auto* t_data = w->data<T>();
    auto* bottom_l_trans_data = tmp->mutable_data<T>(ctx.GetPlace());
    memset(bottom_l_trans_data, 0.0,
           tmp->dims()[0] * tmp->dims()[1] * sizeof(T));

    auto blas = pten::funcs::GetBlas<platform::CPUDeviceContext, T>(ctx);

    call_gemm(blas, CblasNoTrans, CblasNoTrans, x->dims()[0], dim_t * dim_in,
              dim_in, 1.0f, bottom_l_data, t_data, 0.0f, bottom_l_trans_data);

    for (size_t b = 0; b < x->lod()[0].size() - 1; b++) {
      for (int t = 0; t < dim_t; t++) {
        size_t len_l = offset_l[b + 1] - offset_l[b];
        size_t len_r = offset_r[b + 1] - offset_r[b];
        auto* top_data = out_data + top_offset[b] + t * len_l * len_r;
        const auto* l_t_data =
            bottom_l_trans_data + offset_l[b] * dim_t * dim_in + t * dim_in;
        const auto* r_data = bottom_r_data + offset_r[b] * dim_in;
        auto blas_2 = pten::funcs::GetBlas<platform::CPUDeviceContext, T>(ctx);
        call_gemm_with_lda(blas_2, CblasNoTrans, CblasTrans, len_l, len_r,
                           dim_in, 1.0f, l_t_data, r_data, 0.0f, top_data,
                           dim_t * dim_in);
      }
    }

    framework::LoD out_lod;
    out_lod.push_back(top_offset);

    out->set_lod(out_lod);
  }
};

template <typename DeviceContext, typename T>
class CPUMatchMatrixTensorOPGradKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
    auto* x = ctx.Input<LoDTensor>("X");
    auto* y = ctx.Input<LoDTensor>("Y");
    auto* w = ctx.Input<Tensor>("W");
    auto* tmp = ctx.Input<LoDTensor>("Tmp");

    int dim_t = ctx.Attr<int>("dim_t");
    int64_t dim_in = x->dims()[1];

    const auto& offset_l = x->lod()[0];
    const auto& offset_r = y->lod()[0];
    std::vector<size_t> top_offset;
    size_t top_size = 0;
    top_offset.push_back(top_size);
    for (size_t b = 0; b < x->lod()[0].size() - 1; b++) {
      size_t len_l = offset_l[b + 1] - offset_l[b];
      size_t len_r = offset_r[b + 1] - offset_r[b];
      top_size += dim_t * len_l * len_r;
      top_offset.push_back(top_size);
    }

    auto* bottom_l_data = x->data<T>();
    auto* bottom_r_data = y->data<T>();
    auto* bottom_l_trans_data = tmp->data<T>();

    auto* d_out = ctx.Input<LoDTensor>(framework::GradVarName("Out"));
    auto* d_x = ctx.Output<LoDTensor>(framework::GradVarName("X"));
    auto* d_y = ctx.Output<LoDTensor>(framework::GradVarName("Y"));

    Tensor tmp_grad;
    tmp_grad.Resize(tmp->dims());
    auto* d_tmp_data = tmp_grad.mutable_data<T>(ctx.GetPlace());
    auto* top_diff = d_out->data<T>();
    auto* bottom_l_diff = d_x->mutable_data<T>(ctx.GetPlace());
    auto* bottom_r_diff = d_y->mutable_data<T>(ctx.GetPlace());
    auto* bottom_l_trans_diff = const_cast<T*>(d_tmp_data);
    memset(bottom_l_diff, 0.0, x->dims()[0] * x->dims()[1] * sizeof(T));
    memset(bottom_r_diff, 0.0, y->dims()[0] * y->dims()[1] * sizeof(T));
    memset(bottom_l_trans_diff, 0.0,
           tmp->dims()[0] * tmp->dims()[1] * sizeof(T));

    for (size_t b = 0; b < x->lod()[0].size() - 1; b++) {
      for (int t = 0; t < dim_t; t++) {
        size_t len_l = offset_l[b + 1] - offset_l[b];
        size_t len_r = offset_r[b + 1] - offset_r[b];

        for (size_t i = 0; i < len_l; i++) {
          for (size_t j = 0; j < len_r; j++) {
            auto diff =
                top_diff[top_offset[b] + t * len_l * len_r + i * len_r + j];
            auto* l_trans_data = bottom_l_trans_data +
                                 (offset_l[b] + i) * dim_in * dim_t +
                                 t * dim_in;
            auto* l_trans_diff = bottom_l_trans_diff +
                                 (offset_l[b] + i) * dim_in * dim_t +
                                 t * dim_in;
            auto* r_data = bottom_r_data + (offset_r[b] + j) * dim_in;
            auto* r_diff = bottom_r_diff + (offset_r[b] + j) * dim_in;
            if (diff != 0.0) {
              axpy(r_data, l_trans_diff, dim_in, diff);
              axpy(l_trans_data, r_diff, dim_in, diff);
            }
          }
        }
      }
    }

    auto blas = pten::funcs::GetBlas<platform::CPUDeviceContext, T>(ctx);

    auto* t_data = w->data<T>();
    auto* d_w = ctx.Output<Tensor>(framework::GradVarName("W"));
    auto* t_diff = d_w->mutable_data<T>(ctx.GetPlace());
    memset(t_diff, 0.0, w->dims()[0] * w->dims()[1] * w->dims()[2] * sizeof(T));
    // bottom_diff
    call_gemm(blas, CblasNoTrans, CblasTrans, x->dims()[0], dim_in,
              dim_t * dim_in, 1.0f, bottom_l_trans_diff, t_data, 1.0f,
              bottom_l_diff);

    // t_diff
    call_gemm(blas, CblasTrans, CblasNoTrans, dim_in, dim_t * dim_in,
              x->dims()[0], 1.0f, bottom_l_data, bottom_l_trans_diff, 1.0f,
              t_diff);
  }
};

template <typename T>
class MatchMatrixTensorGradOpMaker : public framework::SingleGradOpMaker<T> {
 public:
  using framework::SingleGradOpMaker<T>::SingleGradOpMaker;

 protected:
  void Apply(GradOpPtr<T> grad_op) const override {
    grad_op->SetType("match_matrix_tensor_grad");
    grad_op->SetInput("X", this->Input("X"));
    grad_op->SetInput("Y", this->Input("Y"));
    grad_op->SetInput("W", this->Input("W"));
    grad_op->SetInput("Tmp", this->Output("Tmp"));
    grad_op->SetInput(framework::GradVarName("Out"), this->OutputGrad("Out"));
    grad_op->SetOutput(framework::GradVarName("X"), this->InputGrad("X"));
    grad_op->SetOutput(framework::GradVarName("Y"), this->InputGrad("Y"));
    grad_op->SetOutput(framework::GradVarName("W"), this->InputGrad("W"));
    grad_op->SetAttrMap(this->Attrs());
  }
};

}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
REGISTER_OPERATOR(
    match_matrix_tensor, ops::MatchMatrixTensorOP,
    ops::MatchMatrixTensorOpMaker,
    ops::MatchMatrixTensorGradOpMaker<paddle::framework::OpDesc>,
    ops::MatchMatrixTensorGradOpMaker<paddle::imperative::OpBase>);
REGISTER_OPERATOR(match_matrix_tensor_grad, ops::MatchMatrixTensorOpGrad);

REGISTER_OP_CPU_KERNEL(match_matrix_tensor,
                       ops::CPUMatchMatrixTensorOPKernel<
                           paddle::platform::CPUDeviceContext, float>);

REGISTER_OP_CPU_KERNEL(match_matrix_tensor_grad,
                       ops::CPUMatchMatrixTensorOPGradKernel<
                           paddle::platform::CPUDeviceContext, float>);