#pragma once
#include "caffe2/core/operator.h"
namespace caffe2 {
template <typename T>
struct FtrlParams {
explicit FtrlParams(OperatorBase* op)
: alphaInv(1.0 / op->GetSingleArgument<float>("alpha", 0.005f)),
beta(op->GetSingleArgument<float>("beta", 1.0f)),
lambda1(op->GetSingleArgument<float>("lambda1", 0.001f)),
lambda2(op->GetSingleArgument<float>("lambda2", 0.001f)) {}
T alphaInv;
T beta;
T lambda1;
T lambda2;
};
// TODO(dzhulgakov): implement GPU version if necessary
template <typename T, class Context>
class FtrlOp final : public Operator<Context> {
public:
USE_OPERATOR_CONTEXT_FUNCTIONS;
FtrlOp(const OperatorDef& operator_def, Workspace* ws)
: Operator<Context>(operator_def, ws), params_(this) {
CAFFE_ENFORCE(
!HasArgument("alpha") || ALPHA >= InputSize(),
"Cannot specify alpha by both input and argument");
}
bool RunOnDevice() override;
protected:
FtrlParams<T> params_;
INPUT_TAGS(VAR, N_Z, GRAD, ALPHA);
OUTPUT_TAGS(OUTPUT_VAR, OUTPUT_N_Z);
};
template <typename T>
class SparseFtrlOp final : public Operator<CPUContext> {
public:
SparseFtrlOp(const OperatorDef& operator_def, Workspace* ws)
: Operator<CPUContext>(operator_def, ws), params_(this) {
CAFFE_ENFORCE(
!HasArgument("alpha") || ALPHA >= InputSize(),
"Cannot specify alpha by both input and argument");
}
bool RunOnDevice() override {
// run time learning rate override
if (ALPHA < InputSize()) {
CAFFE_ENFORCE_EQ(Input(ALPHA).numel(), 1, "alpha should be real-valued");
params_.alphaInv = 1.0 / *(Input(ALPHA).template data<T>());
}
// Use run-time polymorphism
auto& indices = Input(INDICES);
if (indices.template IsType<int32_t>()) {
DoRun<int32_t>();
} else if (indices.template IsType<int64_t>()) {
DoRun<int64_t>();
} else {
LOG(FATAL) << "Unsupported type of INDICES in SparseFtrlOp: "
<< indices.dtype().name();
}
return true;
}
protected:
FtrlParams<T> params_;
INPUT_TAGS(VAR, N_Z, INDICES, GRAD, ALPHA);
OUTPUT_TAGS(OUTPUT_VAR, OUTPUT_N_Z);
private:
template <typename SIndex>
void DoRun();
};
}