adam_op.cc 5.3 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
/* 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. */

#include "paddle/operators/adam_op.h"

namespace paddle {
namespace operators {

class AdamOp : public framework::OperatorWithKernel {
 public:
  using framework::OperatorWithKernel::OperatorWithKernel;

  void InferShape(framework::InferShapeContext *ctx) const override {
    PADDLE_ENFORCE(ctx->HasInput("Param"),
                   "Input(Param) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("Grad"),
                   "Input(Grad) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("Moment1"),
                   "Input(Moment1) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("Moment2"),
                   "Input(Moment2) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("LearningRate"),
                   "Input(LearningRate) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("Beta1Pow"),
                   "Input(Beta1Pow) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasInput("Beta2Pow"),
                   "Input(Beta2Pow) of AdamOp should not be null.");

    PADDLE_ENFORCE(ctx->HasOutput("ParamOut"),
                   "Output(ParamOut) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasOutput("Moment1Out"),
                   "Output(Moment1Out) of AdamOp should not be null.");
    PADDLE_ENFORCE(ctx->HasOutput("Moment2Out"),
                   "Output(Moment2Out) of AdamOp should not be null.");

    auto lr_dims = ctx->GetInputDim("LearningRate");
    PADDLE_ENFORCE_EQ(framework::product(lr_dims), 1,
                      "Learning rate should have 1 dimension");
    auto beta1_pow_dims = ctx->GetInputDim("Beta1Pow");
    PADDLE_ENFORCE_EQ(framework::product(beta1_pow_dims), 1,
                      "Beta1 power accumulator should have 1 dimension");
    auto beta2_pow_dims = ctx->GetInputDim("Beta2Pow");
54 55
    PADDLE_ENFORCE_EQ(framework::product(beta2_pow_dims), 1,
                      "Beta2 power accumulator should have 1 dimension");
56 57 58 59 60 61 62

    auto param_dims = ctx->GetInputDim("Param");
    PADDLE_ENFORCE_EQ(
        param_dims, ctx->GetInputDim("Grad"),
        "Param and Grad input of AdamOp should have same dimension");
    PADDLE_ENFORCE_EQ(
        param_dims, ctx->GetInputDim("Moment1"),
63
        "Param and Moment1 input of AdamOp should have same dimension");
64 65
    PADDLE_ENFORCE_EQ(
        param_dims, ctx->GetInputDim("Moment2"),
66
        "Param and Moment2 input of AdamOp should have same dimension");
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105

    ctx->SetOutputDim("ParamOut", param_dims);
    ctx->SetOutputDim("Moment1Out", param_dims);
    ctx->SetOutputDim("Moment2Out", param_dims);
  }
};

class AdamOpMaker : public framework::OpProtoAndCheckerMaker {
 public:
  AdamOpMaker(framework::OpProto *proto, framework::OpAttrChecker *op_checker)
      : OpProtoAndCheckerMaker(proto, op_checker) {
    AddInput("Param", "(Tensor) Input parameter");
    AddInput("Grad", "(Tensor) Input gradient");
    AddInput("LearningRate", "(Tensor) Learning rate");
    AddInput("Moment1", "(Tensor) Input first moment");
    AddInput("Moment2", "(Tensor) Input second moment");
    AddInput("Beta1Pow", "(Tensor) Input beta1 power accumulator");
    AddInput("Beta2Pow", "(Tensor) Input beta2 power accumulator");

    AddOutput("ParamOut", "(Tensor) Output parameter");
    AddOutput("Moment1Out", "(Tensor) Output first moment");
    AddOutput("Moment2Out", "(Tensor) Output second moment");

    AddAttr<float>("beta1",
                   "(float, default 0.9) "
                   "Exponential decay rate for the "
                   "first moment estimates.")
        .SetDefault(0.9f);
    AddAttr<float>("beta2",
                   "(float, default 0.999) "
                   "exponential decay rate for the "
                   "second moment estimates.")
        .SetDefault(0.999f);
    AddAttr<float>("epsilon",
                   "(float, default 1.0e-8) "
                   "Constant for numerical stability")
        .SetDefault(1.0e-8f);

    AddComment(R"DOC(
106
Adam Optimizer.
107 108

This implements the Adam optimizer from Section 2 of the Adam
109 110 111
paper : https://arxiv.org/abs/1412.6980.
Adam is a first-order gradient-based optimization method based on
adaptive estimates of lower-order moments.
112 113 114

Adam updates:

115 116 117 118 119
$$moment_1_{out} = \beta_1 * moment_1 + (1 - \beta_1) * grad \break
moment_2_{out} = \beta_2 * moment_2 + (1 - \beta_2) * grad * grad \break
learningRate = learningRate *
                  $\sqrt{(1 - \beta_2_{pow})}$ / (1 - \beta_1_{pow}) \break
paramOut = param - learningRate * moment_1/ ($\sqrt{(moment_2)} + \epsilon)$$
120 121 122 123 124 125 126 127 128 129 130

)DOC");
  }
};
}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
REGISTER_OP_WITHOUT_GRADIENT(adam, ops::AdamOp, ops::AdamOpMaker);
REGISTER_OP_CPU_KERNEL(adam,
                       ops::AdamOpKernel<paddle::platform::CPUPlace, float>);
反馈
建议
客服 返回
顶部