Merge pull request #16518 from zhoukunsheng/rsqrt

Rsqrt

paddle/fluid/API.spec

@@ -324,6 +324,7 @@ paddle.fluid.layers.atan (ArgSpec(args=['x', 'name'], varargs=None, keywords=Non
 paddle.fluid.layers.tanh_shrink (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '1e521554b9fdda9061ec6d306f0709b7'))
 paddle.fluid.layers.softshrink (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '9eef31597bbafa2bd49691e072296e13'))
 paddle.fluid.layers.sqrt (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', 'e9e27491c39ac74d0b1ffe506aec0ebb'))
+paddle.fluid.layers.rsqrt (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', 'c445467ebe58b3c0d7f0bba7795b6f56'))
 paddle.fluid.layers.abs (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '64650ac42cf82e9920cb0b172b1d29fd'))
 paddle.fluid.layers.ceil (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', 'c75d67dc5fe28f68e4cfffead4f698ad'))
 paddle.fluid.layers.floor (ArgSpec(args=['x', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '647b16c5da5ef909649ae02abb434973'))

paddle/fluid/operators/activation_op.cc

@@ -227,6 +227,15 @@ $out = \sqrt{x}$
 
 )DOC";
 
+UNUSED constexpr char RsqrtDoc[] = R"DOC(
+Rsqrt Activation Operator.
+
+Please make sure input is legal in case of numeric errors.
+
+$out = \frac{1}{\sqrt{x}}$
+
+)DOC";
+
 UNUSED constexpr char AbsDoc[] = R"DOC(
 Abs Activation Operator.
 
@@ -575,6 +584,7 @@ REGISTER_ACTIVATION_OP_MAKER(Gelu, GeluDoc);
 REGISTER_ACTIVATION_OP_MAKER(Tanh, TanhDoc);
 REGISTER_ACTIVATION_OP_MAKER(TanhShrink, TanhShrinkDoc);
 REGISTER_ACTIVATION_OP_MAKER(Sqrt, SqrtDoc);
+REGISTER_ACTIVATION_OP_MAKER(Rsqrt, RsqrtDoc);
 REGISTER_ACTIVATION_OP_MAKER(Abs, AbsDoc);
 REGISTER_ACTIVATION_OP_MAKER(Ceil, CeilDoc);
 REGISTER_ACTIVATION_OP_MAKER(Floor, FloorDoc);
@@ -586,6 +596,7 @@ REGISTER_ACTIVATION_OP_MAKER(Log, LogDoc);
 REGISTER_ACTIVATION_OP_MAKER(Square, SquareDoc);
 REGISTER_ACTIVATION_OP_MAKER(Softplus, SoftplusDoc);
 REGISTER_ACTIVATION_OP_MAKER(Softsign, SoftsignDoc);
+
 }  // namespace operators
 }  // namespace paddle
 

paddle/fluid/operators/activation_op.h

@@ -511,6 +511,26 @@ struct SqrtGradFunctor : public BaseActivationFunctor<T> {
   static constexpr ActBwdOpFwdDeps FwdDeps() { return kDepOut; }
 };
 
+// rsqrt(x) = x^(-1/2)
+template <typename T>
+struct RsqrtFunctor : public BaseActivationFunctor<T> {
+  template <typename Device, typename X, typename Out>
+  void operator()(Device d, X x, Out out) const {
+    out.device(d) = x.rsqrt();
+  }
+};
+
+template <typename T>
+struct RsqrtGradFunctor : public BaseActivationFunctor<T> {
+  template <typename Device, typename X, typename Out, typename dOut,
+            typename dX>
+  void operator()(Device d, X x, Out out, dOut dout, dX dx) const {
+    dx.device(d) = static_cast<T>(-0.5) * dout * out * out * out;
+  }
+
+  static constexpr ActBwdOpFwdDeps FwdDeps() { return kDepOut; }
+};
+
 // ceil(x) = ceiling(x)
 template <typename T>
 struct CeilFunctor : public BaseActivationFunctor<T> {
@@ -1191,6 +1211,7 @@ struct SwishGradFunctor : public BaseActivationFunctor<T> {
   __macro(atan, Atan, AtanFunctor, AtanGradFunctor);                          \
   __macro(softshrink, SoftShrink, SoftShrinkFunctor, SoftShrinkGradFunctor);  \
   __macro(sqrt, Sqrt, SqrtFunctor, SqrtGradFunctor);                          \
+  __macro(rsqrt, Rsqrt, RsqrtFunctor, RsqrtGradFunctor);                      \
   __macro(abs, Abs, AbsFunctor, AbsGradFunctor);                              \
   __macro(ceil, Ceil, CeilFunctor, ZeroGradFunctor);                          \
   __macro(floor, Floor, FloorFunctor, ZeroGradFunctor);                       \

python/paddle/fluid/layers/ops.py

@@ -27,6 +27,7 @@ __activations_noattr__ = [
     'tanh_shrink',
     'softshrink',
     'sqrt',
+    'rsqrt',
     'abs',
     'ceil',
     'floor',

python/paddle/fluid/tests/unittests/test_activation_op.py

@@ -192,6 +192,23 @@ class TestSqrt(TestActivation):
         self.check_grad(['X'], 'Out', max_relative_error=0.007)
 
 
+class TestRsqrt(TestActivation):
+    def setUp(self):
+        self.op_type = "rsqrt"
+        self.init_dtype()
+
+        x = np.random.uniform(0.1, 1, [2, 3]).astype(self.dtype)
+        out = 1.0 / np.sqrt(x)
+
+        self.inputs = {'X': OpTest.np_dtype_to_fluid_dtype(x)}
+        self.outputs = {'Out': out}
+
+    def test_check_grad(self):
+        if self.dtype == np.float16:
+            return
+        self.check_grad(['X'], 'Out', max_relative_error=0.0005)
+
+
 class TestAbs(TestActivation):
     def setUp(self):
         self.op_type = "abs"