inline void RMSPropOptimizer::CreateState_(int index, NDArray weight) {
n_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*n_[index] = 0;
g_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*g_[index] = 0;
delta_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*delta_[index] = 0;
}
inline void SGDOptimizer::CreateState_(int index, NDArray weight) {
if (params_.count("momentum") == 0) {
states_[index] = nullptr;
} else {
states_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*states_[index] = 0;
}
}
inline void Monitor::toc_print() {
auto results = toc();
std::vector<float> data(1);
for (auto& stat : results) {
NDArray ndarray = std::get<2>(stat);
std::string str;
if (ndarray.Size() == 1) {
if (ndarray.GetContext().GetDeviceType() != DeviceType::kGPU) {
data[0] = ndarray.GetData()[0];
} else {
ndarray.SyncCopyToCPU(&data);
}
str = std::to_string(data[0]);
} else {
std::ostringstream out;
out << ndarray;
str = out.str();
}
LG << "Batch: " << std::get<0>(stat) << ' ' << std::get<1>(stat) << ' ' << str;
}
}
inline void AdaDeltaOptimizer::CreateState_(int index, NDArray weight) {
acc_g_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*acc_g_[index] = 0;
acc_delta_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*acc_delta_[index] = 0;
}
inline void AdaGradOptimizer::CreateState_(int index, NDArray weight) {
history_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*history_[index] = 0;
}
inline void AdamOptimizer::CreateState_(int index, NDArray weight) {
mean_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*mean_[index] = 0;
var_[index] = new NDArray(weight.GetShape(), weight.GetContext());
*var_[index] = 0;
}
