init

paddle/operators/adagrad_op.h

@@ -47,8 +47,8 @@ class AdagradOpKernel : public framework::OpKernel<T> {
           *ctx.Input<framework::Tensor>("Grad"));
       auto moment = framework::EigenVector<T>::Flatten(
           *ctx.Input<framework::Tensor>("Moment"));
-      auto lr = framework::EigenVector<T>::Flatten(
-          *ctx.Input<framework::Tensor>("LearningRate"));
+      auto* learning_rate = ctx.Input<framework::Tensor>("LearningRate");
+      auto* lr = learning_rate->data<T>();
 
       auto param_out = framework::EigenVector<T>::Flatten(*param_out_tensor);
       auto moment_out = framework::EigenVector<T>::Flatten(*moment_out_tensor);
@@ -57,7 +57,7 @@ class AdagradOpKernel : public framework::OpKernel<T> {
       moment_out.device(*place) = moment + grad * grad;
       Eigen::DSizes<int, 1> m_dsize(moment_out_tensor->numel());
       param_out.device(*place) =
-          param - lr.broadcast(m_dsize) * grad / (moment_out.sqrt() + epsilon);
+          param - lr[0] * grad / (moment_out.sqrt() + epsilon);
     } else if (grad_var->IsType<framework::SelectedRows>()) {
       auto* param_tensor = ctx.Input<framework::Tensor>("Param");
       PADDLE_ENFORCE_EQ(param_tensor, param_out_tensor);