crossentropy grad op (#3186)

* init cross entropy graident

* add crossentropy grad op

* remove details

* fix static compile

paddle/operators/cross_entropy_op.cc

@@ -36,6 +36,17 @@ class OnehotCrossEntropyOp : public OperatorWithKernel {
   }
 };
 
+class OnehotCrossEntropyGradientOp : public OperatorWithKernel {
+ protected:
+  void InferShape(const InferShapeContext &ctx) const override {
+    auto X_grad = ctx.Output<Tensor>(framework::GradVarName("X"));
+    auto X = ctx.Input<Tensor>("X");
+
+    // TODO(superjom) add enforce here after helper functions ready
+    X_grad->Resize(X->dims());
+  }
+};
+
 class OnehotCrossEntropyOpMaker : public OpProtoAndCheckerMaker {
  public:
   OnehotCrossEntropyOpMaker(OpProto *proto, OpAttrChecker *op_checker)
@@ -58,3 +69,7 @@ REGISTER_OP(onehot_cross_entropy, ops::OnehotCrossEntropyOp,
             ops::OnehotCrossEntropyOpMaker);
 REGISTER_OP_CPU_KERNEL(onehot_cross_entropy,
                        ops::OnehotCrossEntropyOpKernel<ops::CPUPlace, float>);
+
+REGISTER_OP_CPU_KERNEL(
+    onehot_cross_entropy_grad,
+    ops::OnehotCrossEntropyGradientOpKernel<ops::CPUPlace, float>);

paddle/operators/cross_entropy_op.h

@@ -18,28 +18,53 @@ limitations under the License. */
 namespace paddle {
 namespace operators {
 
+static const float kCrossEntropyLogThreshold{1e-20};
+
 template <typename Place, typename T>
 class OnehotCrossEntropyOpKernel : public OpKernel {
  public:
-  constexpr T LOG_THRESHOLD() const { return static_cast<T>(1e-20); }
-
   void Compute(const ExecutionContext& ctx) const override {
-    auto X = ctx.Input<Tensor>(0);
-    const T* X_data = X->data<T>();
+    auto X = ctx.Input<Tensor>("X");
+    const T* Xdata = X->data<T>();
     const int* label_data = ctx.Input<Tensor>(1)->data<int>();
-    auto Y = ctx.Output<Tensor>(0);
+    auto Y = ctx.Output<Tensor>("Y");
 
     Y->mutable_data<T>(ctx.GetPlace());
 
-    T* Y_data = Y->data<T>();
+    T* Ydata = Y->data<T>();
 
     int batch_size = X->dims()[0];
     int class_num = X->dims()[1];
 
     // Y[i] = -log(X[i][j])
     for (int i = 0; i < batch_size; ++i) {
-      Y_data[i] = -std::log(
-          std::max(X_data[i * class_num + label_data[i]], LOG_THRESHOLD()));
+      Ydata[i] = -std::log(std::max(Xdata[i * class_num + label_data[i]],
+                                    kCrossEntropyLogThreshold));
+    }
+  }
+};
+
+template <typename Place, typename T>
+class OnehotCrossEntropyGradientOpKernel : public OpKernel {
+ public:
+  void Compute(const ExecutionContext& ctx) const override {
+    auto X = ctx.Input<Tensor>("X");
+    auto dX = ctx.Output<Tensor>(framework::GradVarName("X"));
+    auto dY = ctx.Input<Tensor>(framework::GradVarName("Y"));
+    auto label = ctx.Input<Tensor>("label");
+
+    auto* dXdata = dX->template mutable_data<T>(ctx.GetPlace());
+    auto* dYdata = dY->template data<T>();
+    auto* Xdata = X->template data<T>();
+    auto* label_data = label->data<int>();
+
+    const int batch_size = X->dims()[0];
+    const int class_num = X->dims()[1];
+
+    for (int i = 0; i < batch_size; ++i) {
+      dXdata[i * class_num + label_data[i]] =
+          -dYdata[i] / std::max(Xdata[i * class_num + label_data[i]],
+                                kCrossEntropyLogThreshold);
     }
   }
 };

python/paddle/v2/framework/tests/test_cross_entropy_op.py

@@ -18,5 +18,7 @@ class TestSGD(unittest.TestCase):
         self.Y = numpy.array(Y).astype("float32")
 
 
+# TODO(superjom) add gradient check
+
 if __name__ == "__main__":
     unittest.main()