Merge pull request #498 from pinpom/pow

SINGA-475 add Pow operator

Author: nudles

python/singa/autograd.py

@@ -1829,6 +1829,27 @@ def backward(self, dy):
 def leakyrelu(x, a=0.01):
     return LeakyRelu(a)(x)[0]
 
+  
+class Pow(Operation):
+    def __init__(self):
+        super(Pow, self).__init__()
+
+    def forward(self, a, b):
+        if training:
+            self.input = (a, b)
+        return singa.Pow(a, b)
+
+    def backward(self, dy):
+        da1=singa.__mul__(self.input[1], singa.Pow(self.input[0], singa.SubFloat(self.input[1],1.0)))
+        da=singa.__mul__(da1, dy)
+
+        db1=singa.__mul__(singa.Pow(self.input[0],self.input[1]), singa.Log(self.input[0]))
+        db=singa.__mul__(db1, dy)
+
+        return da, db
+
+def pow(a, b):
+    return Pow()(a,b)[0]
 
 class SoftSign(Operation):
     def __init__(self):
@@ -1860,10 +1881,12 @@ def forward(self, x):
         if training:
             self.input = x
         return singa.Sqrt(x)
-
+      
     def backward(self, dy):
         dx = singa.PowFloat(self.input,-0.5)
         dx = singa.MultFloat(dx,0.5)
+        dx = singa.__mul__(dy, dx)
+        return dx
 
 def sqrt(x):
     return Sqrt()(x)[0]

test/python/test_operation.py

@@ -814,6 +814,7 @@ def test_Sub_gpu(self):
         x1.to_device(gpu_dev)
         dy.to_device(gpu_dev)
 
+
         result = autograd.sub(x0, x1)
         dx0, dx1 = result.creator.backward(dy.data)
         DX0 = np.multiply(DY, 1.0)
@@ -822,6 +823,56 @@ def test_Sub_gpu(self):
         np.testing.assert_array_almost_equal(tensor.to_numpy(result), XT, decimal=5)
         np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx0)), DX0, decimal=5)
         np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx1)), DX1, decimal=5)
+        
+    def test_Pow_cpu(self):
+        X0 = np.array([7, 5, 0.2, 0.1, 0.3, 4]).reshape(3, 2).astype(np.float32)
+        X1 = np.array([-1.0, 2.0, -1.0, -2.1, 1.0, -2.0]).reshape(3, 2).astype(np.float32)
+        XT = np.power(X0, X1)
+        
+        DY = np.ones((3, 2), dtype = np.float32)
+        x0 = tensor.from_numpy(X0)
+        x1 = tensor.from_numpy(X1)
+        dy = tensor.from_numpy(DY)
+        x0.to_device(cpu_dev)
+        x1.to_device(cpu_dev)
+        dy.to_device(cpu_dev)
+
+        result = autograd.pow(x0, x1)
+        dx0, dx1 = result.creator.backward(dy.data)
+
+        G0 =  np.multiply(X1, np.power(X0, (X1 - 1.0)) )
+        DX0 = np.multiply(G0, DY)
+        G1 = np.multiply(np.power(X0, X1), np.log(X0) )
+        DX1 = np.multiply(G1, DY)
+
+        np.testing.assert_array_almost_equal(tensor.to_numpy(result), XT, decimal=5)
+        np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx0)), DX0, decimal=4)
+        np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx1)), DX1, decimal=4)
+
+    def test_Pow_gpu(self):
+        X0 = np.array([7, 5, 0.2, 0.1, 0.3, 4]).reshape(3, 2).astype(np.float32)
+        X1 = np.array([-1.0, 2.0, -1.0, -2.1, 1.0, -2.0]).reshape(3, 2).astype(np.float32)
+        XT = np.power(X0, X1)
+        
+        DY = np.ones((3, 2), dtype = np.float32)
+        x0 = tensor.from_numpy(X0)
+        x1 = tensor.from_numpy(X1)
+        dy = tensor.from_numpy(DY)
+        x0.to_device(gpu_dev)
+        x1.to_device(gpu_dev)
+        dy.to_device(gpu_dev)
+
+        result = autograd.pow(x0, x1)
+        dx0, dx1 = result.creator.backward(dy.data)
+
+        G0 =  np.multiply(X1, np.power(X0, (X1 - 1.0)) )
+        DX0 = np.multiply(G0, DY)
+        G1 = np.multiply(np.power(X0, X1), np.log(X0) )
+        DX1 = np.multiply(G1, DY)
+
+        np.testing.assert_array_almost_equal(tensor.to_numpy(result), XT, decimal=5)
+        np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx0)), DX0, decimal=4)
+        np.testing.assert_array_almost_equal(tensor.to_numpy(tensor.from_raw_tensor(dx1)), DX1, decimal=4)
 
     def test_SoftSign_cpu(self):
         # y = x / (1 + np.abs(x))