hierarchical_sigmoid_op.h

/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.

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. */

#pragma once
#include <iostream>
#include <vector>
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/framework/selected_rows.h"
#include "paddle/fluid/operators/clip_op.h"
#include "paddle/fluid/operators/math/math_function.h"
#include "paddle/fluid/operators/math/matrix_bit_code.h"
#include "paddle/fluid/platform/transform.h"
namespace paddle {
namespace operators {

template <typename T, int MajorType = Eigen::RowMajor,
          typename IndexType = Eigen::DenseIndex>
using EigenMatrix = framework::EigenMatrix<T, MajorType, IndexType>;
using platform::Transform;

template <typename DeviceContext, typename T>
class HierarchicalSigmoidOpKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
    auto* in = ctx.Input<framework::Tensor>("X");
    auto* w = ctx.Input<framework::Tensor>("W");
    auto* path = ctx.Input<framework::Tensor>("PTable");
    auto* code = ctx.Input<framework::Tensor>("PCode");
    auto* label = ctx.Input<framework::Tensor>("Label");
    auto* bias = ctx.Input<framework::Tensor>("Bias");
    auto* out = ctx.Output<framework::Tensor>("Out");
    auto* pre_out = ctx.Output<framework::Tensor>("PreOut");
    size_t num_classes = static_cast<size_t>(ctx.Attr<int>("num_classes"));
    bool is_custom = false;
    if (path) {
      is_custom = true;
    } else {
      is_custom = false;
    }
    int64_t code_length =
        path ? path->dims()[1] : math::FindLastSet(num_classes - 1);
    int64_t batch_size = in->dims()[0];
    framework::Tensor sum;
    auto& dev_ctx = ctx.template device_context<DeviceContext>();
    auto* pre_out_data = pre_out->mutable_data<T>(
        framework::make_ddim({batch_size, code_length}), ctx.GetPlace());
    auto pre_out_mat = EigenMatrix<T>::From(*pre_out);
    // Not all class(leaf) nodes' path lengths equal code_length, thus init as
    // 0s can avoid out of path's loss.
    math::SetConstant<DeviceContext, T> zero;
    zero(dev_ctx, pre_out, static_cast<T>(0.0));
    auto& place = *ctx.template device_context<DeviceContext>().eigen_device();
    math::RowwiseSum<DeviceContext, T> row_sum;

    std::unique_ptr<math::MatrixBitCodeFunctor<T>> bit_code;
    if (!is_custom) {
      bit_code.reset(new math::MatrixBitCodeFunctor<T>(num_classes,
                                                       label->data<int64_t>()));
    } else {
      bit_code.reset(new math::MatrixBitCodeFunctor<T>(path, code,
                                                       label->data<int64_t>()));
    }

    std::vector<int64_t> sum_dims({batch_size, 1UL});
    sum.mutable_data<T>(framework::make_ddim(sum_dims), ctx.GetPlace());
    auto sum_mat = EigenMatrix<T>::From(sum);
    out->mutable_data<T>(ctx.GetPlace());
    auto out_mat = framework::EigenVector<T>::Flatten(*out);
    if (bias) {
      bit_code->Add(pre_out, *bias);
    }
    bit_code->Mul(pre_out, *w, *in);
    // clip to [-40, 40]
    Transform<DeviceContext> trans;
    trans(ctx.template device_context<DeviceContext>(), pre_out_data,
          pre_out_data + pre_out->numel(), pre_out_data,
          ClipFunctor<T>(static_cast<T>(-40.0), static_cast<T>(40.0)));
    pre_out_mat = -1 * pre_out_mat;
    bit_code->Sum(*pre_out, out, static_cast<T>(-1));
    // use softrelu to calculate cross entropy
    pre_out_mat.device(place) = (static_cast<T>(1.0) + pre_out_mat.exp()).log();
    row_sum(dev_ctx, *pre_out, &sum);
    // TODO(guosheng): Subtract the out of path's loss, since not all
    // class(leaf) nodes' path lengths equal code_length. But it won't break the
    // gradient check since both have the out of path's loss and will cancel out
    // each other.
    out_mat.device(place) = sum_mat + out_mat;
  }
};

template <typename DeviceContext, typename T>
class HierarchicalSigmoidGradOpKernel : public framework::OpKernel<T> {
 public:
  void Compute(const framework::ExecutionContext& ctx) const override {
    auto* in = ctx.Input<framework::Tensor>("X");
    auto* w = ctx.Input<framework::Tensor>("W");
    auto* path = ctx.Input<framework::Tensor>("PTable");
    auto* code = ctx.Input<framework::Tensor>("PCode");
    auto* in_grad = ctx.Output<framework::Tensor>(framework::GradVarName("X"));
    auto* w_grad = ctx.Output<framework::Tensor>(framework::GradVarName("W"));
    auto* bias_grad =
        ctx.Output<framework::Tensor>(framework::GradVarName("Bias"));
    auto* label = ctx.Input<framework::Tensor>("Label");
    auto* pre_out = ctx.Input<framework::Tensor>("PreOut");
    auto* out_grad =
        ctx.Input<framework::Tensor>(framework::GradVarName("Out"));
    framework::Tensor pre_out_grad;

    pre_out_grad.mutable_data<T>(pre_out->dims(), ctx.GetPlace());
    in_grad->mutable_data<T>(ctx.GetPlace());
    w_grad->mutable_data<T>(ctx.GetPlace());
    auto& dev_ctx = ctx.template device_context<DeviceContext>();
    math::SetConstant<DeviceContext, T> zero;
    zero(dev_ctx, in_grad, static_cast<T>(0.0));
    zero(dev_ctx, w_grad, static_cast<T>(0.0));

    size_t num_classes = static_cast<size_t>(ctx.Attr<int>("num_classes"));

    bool is_custom = false;
    if (path) {
      is_custom = true;
    } else {
      is_custom = false;
    }

    std::unique_ptr<math::MatrixBitCodeFunctor<T>> bit_code;
    if (!is_custom) {
      bit_code.reset(new math::MatrixBitCodeFunctor<T>(num_classes,
                                                       label->data<int64_t>()));
    } else {
      bit_code.reset(new math::MatrixBitCodeFunctor<T>(path, code,
                                                       label->data<int64_t>()));
    }

    auto& place = *ctx.template device_context<DeviceContext>().eigen_device();
    auto pre_out_mat = EigenMatrix<T>::From(*pre_out);
    auto pre_out_grad_mat = EigenMatrix<T>::From(pre_out_grad);
    auto out_grad_mat = EigenMatrix<T>::From(*out_grad);

    Eigen::array<int, 2> bcast({{1, static_cast<int>(pre_out_grad.dims()[1])}});

    // softrelu derivative
    pre_out_grad_mat.device(place) =
        static_cast<T>(1.0) - static_cast<T>(1.0) / pre_out_mat.exp();
    bit_code->Sub(&pre_out_grad);  // the gradient of clip(w * x + b)
    pre_out_grad_mat.device(place) =
        pre_out_grad_mat * out_grad_mat.broadcast(bcast);
    // TODO(guosheng): multiply pre_out_grad with subgradient of clipping to
    // be consistent with the clipping in forward.
    if (bias_grad) {
      bias_grad->mutable_data<T>(ctx.GetPlace());
      zero(dev_ctx, bias_grad, static_cast<T>(0.0));
      bit_code->AddGrad(pre_out_grad, bias_grad);
      auto bias_grad_mat = EigenMatrix<T>::From(*bias_grad);
      bias_grad_mat = -1 * bias_grad_mat;
    }
    bit_code->MulGradWeight(pre_out_grad, w_grad, *in);
    bit_code->MulGradError(pre_out_grad, *w, in_grad);
    auto w_grad_mat = EigenMatrix<T>::From(*w_grad);
    auto in_grad_mat = EigenMatrix<T>::From(*in_grad);

    w_grad_mat = -1 * w_grad_mat;
    in_grad_mat = -1 * in_grad_mat;
  }
};

}  // namespace operators
}  // namespace paddle